,^ -CTi

S^^CNI

^ —

. O

-o

5 CD

C^

cr>

^.'ia^zt:::^>~fir>yi^\

Digitized by the Internet Archive

in 2008 with funding from

IVIicrosoft Corporation

http://www.archive.org/details/ethnographycasteOObainuoft

l^ko

^^j^jgp^

Encyclopedia of Indo-Aryan Research

FOUNDED BY (;. BUllLER, CONTINUED BY F. KIEI.HORN,

EDITED BY H. LUDERS AND J. VVACKERNAGEL.

VOL. II, PART 5.

ETHNOGRAPHY

(CASTES AND TRIBES)

SIR ATHELSTANE BAINES

WITH A LIST OF THE MORE IMPORTANT WORKS ON INDIAN ETHNOGRAPHY
BY W. SIEGLING.

^lubcr tl]C |Jntronagc nf Jijis Majesty's |Irt:trifial ^ccrctarg of ^iate for ^nlita.

STRASSBURG

KARL J. TRUBNER
1912.

PS

M. DiiMoiit Schmiberg, StraClmrg.

Encyclopedia of Indo-Aryan Research

FOUNDED BY U. BUHLER, CONTINUED BY F. KIELHORN,
EDITED BY H. LUDERS AND J. WACKERNAGEL.

VOL. II, PART 5.

ETHNOGRAPHY

(CASTES AND TRIBES)

BY

SIR ATHELSTANE BAINES.

INTRODUCTION.

§ I. The subject with which it is proposed to deal in the present
work is that branch of Indian ethnography which is concerned with the
social organisation of the population, or the dispersal of the latter into
definite groups based upon considerations of race, tribe, blood or oc-
cupation. In the main, it takes the form of a descriptive survey of the
return of castes and tribes obtained through the Census of 1901. The
scope of the review, however, is limited to the population of India properly
so called, and does not, therefore, include Burma or the outlying tracts
of Baluchistan, Aden and the Andamans, by the omission of which the
population dealt with is reduced from 294 to 283 millions.

§ 2. It should be borne in mind from the outset, that but for the
fact that this vast aggregate is spread over a continuous area between
Cape Comorin and the Himalaya, and is politically under one rule, the
population does not contain, as a whole, any of the essential elements of
Nationality. Irrespective of racial differences, which, for reasons which will
appear below, are to a great extent outside the Census inquiry, the
Language, falling under no less than 147 heads, varies from Province to
Province, each of the principal tongues having its dialects whose Shibboleth
infallibly denotes the stranger a hundred miles or so from his native village.
Society, again, is split up into almost innumerable self-contained divisions,
under sacerdotal prohibition from intermarriage and domestic intercourse
with each other. Religion, moreover, constitutes a well-defined distinction
only in the case of creeds introduced from abroad, and the Faith returned
under a single title, itself of foreign origin, by nearly three fourths of the
population covers a vast and incoherent collection of beliefs and forms of
worship, from the tribal animism of the primitive denizens of the forest to
those involving the most refined metaphysical conceptions. Neither religion
nor language, then will be here discussed more than cursorily, and solely
in their bearings upon the ethnography of the country. Full information
upon the philology and the main currents of religious belief of India will
be found in special treatises upon those subjects in other volumes of this
Encyclopaedia. Moreover, neither creed nor mother-tongue affords an
adequate, or even an approximate indication of the great fundamental
variety of race, a subject which also escapes the Census inquiry since

Indo-Aryan Research. II. 5. 1

5. Ethnography.

the latter takes cognisance, perforce, of existing facts only, whilst race
has been for centuries obscured by the operation of the two most pre-
valent forms of religious profession. The plastic and assimilative nature
of Brahmanism absorbs, whilst the uncompromising tendencies of Islam
obliterate, distinctions of race equally with those of doctrine and cere-
monial, and both have their effect in diminishing the popularity of the
more restricted vernaculars. The veil of superficial uniformity which has
thus been drawn over the actual elements from which Indian society has
been formed can only be removed, and then but iwrtially and on con-
jecture perhajis, by recourse to such ethnological evidence as may be
gleaned from tradition and literature, with the aid, in certain directions,
of anthropometrical investigation, so far as it has yet been carried. Purity
of descent is no more a general characteristic of the population of India
than it is of any other old civilisation in the Eastern Hemisphere in which
geographical conformation admits of access from the North. In the Upper,
or Continental, portion of India that purity is probably found in the upper
classes of the Panjab and Rajputana. It exists, too, at the ojiposite end
of the social ladder, amongst the Hill tribes of the Belt dividing the above
portion of India from the Peninsula. South of that barrier, again, the
population, except along parts of the West Coast, is comparatively homo-
geneous, and the main variations noticeable in it are not more marked
than those which may reasonably be attributed to secular differences in
habits and pursuits. The principal physical features of the country have
to be taken into account in connection with its ethnography, as they have
played a highly important part in determining the racial distribution of
the population. To put it briefly, India can only be entered from the
north by any considerable body of men by passes through the outlying
ranges running southwards from the Himalaya in the western extension
of that great system. In early times, no doubt, access was comparatively
easy by routes debouching on the middle and lower Indus, over country
which is now sandy desert, but which was once the abode of a consi-
derable population. Similarly, on the eastern flank of the Himalaya, the
trend of the lower ranges renders it possible for those accustomed to
forest and mountain life to enter, though not in large bodies, the valley
of the Brahmaputra or the eastern Gangetic Delta. Between the mountains
and the next obstacle, the ranges of Central India, lie the vast alluvial
plain of the Ganges and its tributaries and the open plains of the Five
Rivers. The Central Belt, of considerable width in both hill and forest,
though of insignificant height in comparison with the Himalaya, is yet
sufficiently difficult to have proved an effective obstacle in the infancy
of means of communication and of protective government. It also affords
shelter to a considerable population of the wilder tribes, of old the
guardians of the routes through their territory. As in the case of the
Himalaya, however, the flank can be turned on both east and west, as
the hills do not reach either coast, and the narrow strips intervening
between the ranges and the sea consist of fertile and low-lying country-,
presenting little or no difficulty of passage on the East, at all events, to
the great southern plains and the Dekkan plateau. These prominent na-
tural features have now to be coordinated with the ethnology of India,
so far as our knowledge of the latter extends.

§ 3. The basic population of practically the whole country- consists
of a dark, short and broad-nosed race, with wavy, but not woolly, hair.

Introduction. 3

In the present day it is represented by the wild tribes of the Central Belt,
and in a higher state of culture by the population of the southern portions
of the Peninsula. On philological grounds, the people south of the Belt are
distinguished from those further north. The former, known as Dravidian,
seem always to have kept to their present localities, except in a few
cases where tribes have migrated into the Belt within historic times. The
other race, to which the title of Kol or Munda, is generally attached,
is not known south of the forest Belt, in which it is at the present time
concentrated under its distinctive tribal appellations. Formerly, however,
it was spread over the whole of the great plains of Upper India, and,
according to recent philological discoveries, it is akin, at least in language,
to communities now settled on the borders of Assam, and far to the east
of the Bay of Bengal. Some investigators, indeed, spread its former habitat
over a still wider area. In the east and north-east of India, however, its
identity has been obscured, if not obliterated, by the successive immigra-
tions of people of IMongoloidic race from eastern Tibet and the head
waters of the great Chinese rivers, whose main streams of migration have
sought the sea by the valleys of the Irawadi, Salwin and Mekhong. In the
Gangetic plain the type is traceable throughout the population, slightly,
indeed, along the Jamna, but more distinctly as the east is approached,
and almost everywhere more prevalent as the social position is lower.
This graduation is due to miscegenation between the Kol, who, as far as
ethnography is concerned, may be considered the autochthonous inhabitant
of these tracts, and a taller and fairer race, which entered India by the
passes of the North-west or the plains of Balijchistan. More than one such
race are known to history, but in most cases their impact upon India was
sharp but short; not, at any rate, of a character to leave a permanent
impression upon the population. Such, for instance, was the connection
of the iMacedonians with the Panjab. More durable though still in few
cases amounting to settlement or colonisation, were the principalities set
up from time to time in the North-west by scions of the race or races
termed Scythian, of whom more will be said below. The only immigrating
race of practical importance in connection with the present subject, is
that of the Aryas, whose advent and progress are indirectly, and to a
great extent conjecturally, revealed in the collection of their invocations
handed down from perhaps as early as 3000 B C, in the Rgveda and the
sacerdotal literature appended to it at later dates.

§ 4. From these sources it appears that a number of cognate tribes
of northern race and pastoral habits advanced across and along the Indus
into the Panjab, where they settled after dispossessing the dark tribes
in occupation, relegating them to the position of helots in the service of
the new communities. The Vedic Aryas seem to have lost touch in time
with their original country across the snows, and to have developed their
civilisation on lines peculiarly their own. Their progress eastwards from
the Indus was that of expansion rather than of conquest, as the Kol tribes
seem after a time to have offered no serious resistance. The comparatively
easy conditions of life in sub-tropical circumstances, and the immunity
from attack in force from the west, which was secured by their mountain
rampart, combined to soften the northern fibre of the race, and, in course
of time, the supreme influence over the community was transferred from
the chieftain to the priests, under whose auspices society was organised
in a way that secured the absolute supremacy of their own order. The

5. Ethnography.

system thus established was so elastic in the matter of doctrine and
worship, so simple in its demands upon traditional rites and customs, that
without |)ropaganda or formal conversion, it absorbed and continues to
absorb into the pale of orthodoxy the religious and domestic observances
of all the non-Aryan tribes with which it came into contact. As a neces-
sary result, ethnical distinctions are thus obliterated by religious termino-
logy, and, along with the tribal nomenclature, tribal languages have long
tended to disappear from usage. This has been the case throughout the
Gangetic valley, in Central India, and along the northern districts of the
Western coast, in none of which tracts is creed or language an indication
of racial origin. In the first named region, too, the physical characteristics
of the masses denote clearly the admixture of Kol with Aryan blood, a
blend which, as above stated, grows more perceptible as the distance
from the centres of Aryan settlement increases. The striking differences
in this respect between the population of the Panjab and northern Raj-
putana and that east of the Jamna appears to be due both to the stricter
maintenance of the purity of the original northern stock, and also to the
recruitment of that stock through the subsequent occupation of the first-
mentioned tracts by communities from beyond the Himalaya. The most
important of the latter are the various tribes known in ancient Indian
literature by the probably generic title of S'aka, or Scythians, the greater
portion of whom made their way south by way of Bactria. In mure than
one instance the dynasty establishing itself in India lasted so long and
penetrated so far into the interior, that it is almost certain to have left
a physical, as well as a political, impress upon the population. The case
of the Yetha Hunas, or White Huns, is c)ne in point. After the usual
vicissitudes north of the great ranges, they ruled in Central India for a
considerable period, and, long before their overthrow, they seem to have
been absorbed into the local chieftainry of Rajputana and Malva. For
several generations, too, a Pahlava, or Parthian, dynasty held sway on the
lower Indus. The origin of most of these peoples was probably in the
^longoloid regions of north-east Asia, but recent investigators appear to
consider that it is not improbable that at least one, and that an important
dynasty in Northern India, was of .-Vryan race, driven southwards by the
pressure on west-central Asia from the north-east. Whatever the actual
race, the point relevant to the present iiuestion is that they were all
northerners, and thus alien in blood and physique to the prc-.\ryan in-
habitants of India.

§ 5. The Connection of the Aryas with Dravidian India seems to
have been of a different character from that established in the Gangetic
region and the Panjab. There does not appear to have been any coloni-
sation, and little, if any, cross-breeding. It may be fairly conjectured that
the open and fertile plains of the south-east afforded opportunities for
civilisation upon local lines to an extent which, by the time the .\ryas had
spread to the means of access from the north, had placed the Dravidian
communities in a much stronger position than the Kol tribes of the
Continental plains. From the Aryan additions to the vocabulary of the
vernacular tongues and the s])ecial features of the Brahmanism and the
social system of the South it may be inferred that the influence of .\ryan
civilisation was there of a missionary, not political or military, character.
The cloak of Brahmanic orthodoxy was thrown over the local deities and
ceremonial, and social divisions adopted the Brahmanic organisation : but,

Introduction.

beyond the introduction of a certain contingent of Brahmans as teachers
and advisers, no Aryan blood was infused into the population. Along the
western coast, however, which is cut off from the Tamil country and the
Dekkan by the Sahyadri range, tradition assigns a northern origin to
several of the more important communities, and is confirmed by physical
appearance and certain sjiccial customs.

!; 6. It remains to mention the more modern accretions to the peoples
of India received from foreign countries, but now permanently established
in the land of their adoption.

Of movements of this description which have had a racial signi-
ficance, that which took place under the auspices of the followers of
Muhammad first claims attention. It must be noted, however, that, on the
whole, the extent to which it introduced fresh blood into the country is
of far less importance than its religious and political influence. India con-
tains, it is true, more Muslim than any other country in the world, and
votaries of their faith are found in every part of it; but, except in the
territories bordering upon the exclusively Muslim States of Afghanistan
and Balijchistan, the community consists almost entirely of local converts
from Rrahmanism, without any admixture of foreign blood. In Upper India,
colonies of considerable importance were left by successive waves of
invasion, especially in and round the cities founded or occupied by the
conquering races. In the case of the Moghal dynasties, military and ad-
ministrative centres were established far down the Ganges and on the
western coast. The Arabs, too, have been in commercial intercourse with
that coast from time immemorial, and have planted permanent settlements
as far south as Malabar. The largest aggregates, however, of foreign
Muslim are those recruited from the Indus frontier, and settled not far
from that river. The conversion of Sindh and Kashmir has long been
almost complete, and that of the eastern tracts of the great Delta of the
Ganges and Brahmaputra is in active progress, and already extends to
more than half the population. With this exception, the proportion of
Muslim diminishes, like that of the Aryan stock, southwards and eastwards
from the Panjab, and is very small amongst the Dravidians, and scarcely
existent in the Central Belt of hills and forests. From the standpoint of
ethnography it is not to be assumed that the results of conversion to
Islam extend no further than the substitution of one dogma or ritual for
another, as is the case, to a great extent, when a lower race is absorbed
into Brahmanism. The acceptance of the monotheistic creed entails, as
a rule, material expansion of the matrimonial field and of the social horizon
generally, with a wider range of diet also, all of which tend to differentiate,
after a generation or two, the converted community from that to which
it originally belonged, the modification extending to physical as well as
to other attributes.

§ 7. Another community which, as regards the majority of its
members, is the result of apostolic zeal rather than of immigration, is that
of the Christians in India, of whom more than 91 per cent are native to
the country and another 3 per cent of mixed European and native origin.
The remainder are practically sojourners only, and comprise the European
military and civil establishments, the mercantile communities of the larger
cities, and the considerable staff of the railway systems. The conversion
of certain localities, chiefly on the Malabar coast, is alleged to date from
the first century of the Christian era; but until the arrival of the Portu-

6 5- Ethnography.

guesc, the propaganda was not extended far beyond the original settle-
ments of the Nestorian Church. The Roman Catholic missionaries, under
the political aegis of Goa, ranged over a large portion of Southern India,
and, to this day, three fourths of the Christian population of India belongs
to the Dravidian tracts, and more than half to the Church of Rome. The
differentiation of the convert to this religion from his Brahmanic fellows
varies, usually according to the numbers and homogeneity of the local
congregation. The breach with old custom is more marked where con-
version is comparatively sporadic, and slighter in the case where Christianity
has been hereditary for generations, or, if of comparatively recent accep-
tance, has been embraced by considerable numbers of more or less the
same social position. This position, owing mainly to the restrictions of
the caste system, is generally low, as the change is there not only less
of a sacrifice to people who have no hope of rising, but may even bring
with it some chances of ameliorating their lot.

§ 8. At the very opposite pole to the Muslim and Christians in
regard to recruitment by propaganda of their religion, stand the small but
well defined body of Parsis. The original settlers of this race were driven
out of Iran by the Muslim in the 7th century, and the bulk of their
descendants are still to be found in and round the tract upon which they
first landed, on the coast north of Bombay. The opening of the latter by
the British as the commercial emporium of western India, induced many
families of Parsis to migrate thither, and from this centre they have spread
all over the country to such an extent that, though their aggregate nimibers
is only just over 93000, there is scarcely a large town in India in which
a few families of Parsi traders are not resident. From their arrival in the
country the Parsis made a point of keeping their race and ritual unsullied
by intercourse with their neighbours, and to this particularism is due to
some extent, their very slow rate of increase. It is remarkable, however,
that with this strict maintenance of their customs and ritual, and their
abstinence from intermarriage with Indians, the Parsis have long lost all
hold of their original language, Pahlavi, except in their liturgy, and uni-
versally make use of Gujarati as their mother-tongue.

§ 9. In addition to the Christians, Parsis and Arabs, the west coast
of India has also afforded refuge to successive small bodies of Israelites,
of which the more ancient, at all events, hold the tradition that like the
Parsis, they were driven by persecution from their fatherland. Like the
sons of Iran, again, they have kept up their religion and customs and
lost their mother-tongue. The earliest colony is that of Cochin, on the
Malabar coast, which dates from the Christian era, if not from an earlier
period. It consists of two sections, the White, which has kept its breed
pure, and gets its brides occasionally from Syria and Baghdad, and the
Black, which is suspected of intermarriage with Indians or of the incor-
poration of local converts in days of yore, and is therefore socially avoided
by the others. The total number of both communities does not exceed
1300, and is not increasing. Another Jewish settlement of apparently
distinct origin from those further south, is that of the Beni-Israel, on the
mainland near Bombay. The members thereof possess the i^hysical charac-
teristics of their race, and keep up their religious observances, though
they have adopted the dress and language of their Maratha neighbours.
Unlike their compatriots in general, they are engaged chiefly in cultivation,
and have taken to a considerable extent also to military service in the

Introduction.

British Indian army. They have the same tradition as those of Cochin as
to their exile from their country under persecution, but seem to have a
iaxer grip of their past than the latter, and no inclination for alliances
with those of their race beyond the seas. In numbers they greatly surpass
their fellow exiles. The largest community of Jews in India is the com-
paratively recent commercial settlement in Bombay and to a less extent
in Calcutta, of traders from Baghdad, who, whilst permanently settled in
their place of business, keep in close touch with their old home.

§ 10. The above sketch of the ethnological aspect of the subject
will serve to indicate this fact of primary relevance, that, north of the
Dravidian country, the demarcation of race is only ascertainable in the
case of the communities under tribal constitution, such as the Kol of the
Central Belt, the jMongoloid tribes of the North-cast, and the Muslim
immigrants of the North-west. The undoubted racial difference between
the fair people of Rajputana and the Panjab and the masses further east
is obscured, for the purposes of ethnography, by the superstructure of
Brahmanism under which it now lies buried. This survey would be in-
complete, however, without some exposition of the distribution of creed
and language, even though it be restricted to mere numbers. First, then,
in regard to !Mothertongue, it will be seen from Table I given cm the ne.xt
page, that no more than about one person in a thousand returns any language
not peculiar to India or its immediate vicinity, and that one, is probably
a European sojourner. Nine in a thousand speak a frontier dialect, mainly
PashtU, Baluchi, Tibetan or one of the almost innumerable languages of
the hill-tracts between India and Burma. The languages distinguishable as
restricted respectively to special tribes are returned by some 6' \ millions ;
and, on the whole, 96 per cent speak Indo-Aryan languages or Dravidian,
other than those of the hill-tribes. Appended to this volume is a Table
showing the territorial distribution of each of the principal tongues, from
which a conception may be formed of the great linguistic diversity of the
country.

§ II. It will be inferred from what has been stated above, that the
diversity of religion is by no means equal to that of language, so far as
nomenclature is in question. In Table II on the next page, the numbers
of those professing the main forms of belief are given, along with their
relative proportion to the total population.

It must be understood that the term „Tribal Animism" refers to the
religion returned under the tribal name by those who adhere to none of
the wider creeds. Again, the title „Hinduism" is only recognised by the
community to whom it is applied as denoting a distinction between them
and the foreigner. The word was first used by the ?iluslim invaders for
all Indian creeds in which the uncompromising Unitarianism of the follower
of the Prophet detected signs of the worship of idols. It is here taken in
its conventional sense of „the collection of rites, worships, beliefs, traditions
and mythologies that are sanctioned by the sacred books and ordinances
of the Brahmans, and are propagated by Brahmanic teaching" (Lyall). In
practice, this amounts to the application of the title to any Brahmanic
community that has not returned t)ne of the more specific denominations
which can legitimately be included under the general name. Consequently,
the great mass of the people come under it. The prevalence of the different
professions of faith in the principal territorial divisions of India is shown in
a Table appended to this volume.

5- Ethnography.

J

Population

Population returniny

Linguistic Class

!-5i

Languages native to

Total

Per
100,000

Indian
Frontiers

India

Foreign
Counviet

I. Kol-Khervari

lO

3,179,273

1,124

—

3,179,273

II. Dravidian. .

14

56,315,740

19,911

47,943

56,267,797

—

m. Arj-o-Dravidian

*

344,143

122

344,143

—

IV. Indo-Aryan .

20

219,352,079

77,556

54,425

219,297,654

—

V. Iranian. . .

6

1,388,223

491

1,369,133

—

19,090

VI. Tibeto-Burman

62

1,804,776

638

960,585

844,191

—

VII. M6n . . .

2

177,854

63

27

177,827

—

VIII. Tai . . . .

6

3,366

I

3,366

—

IX. Mongolian

4

3,566

I

—

_

3,566

X. Malay . . .

I

26

—

—

—

26

XI. Semitic

3

19,726

7

—

—

19,726

XII. Hamitic

t

185

—

—

—

185

XIII. European. .

23
151

243,109

86

—

—

243,109

Total returned

282,832,066

100,000

2,435,479

280,110,885

285,702

Not returned . .

158,997

—

—

Population . .

—

282,991,063

—

—

—

—

* Gipsy dialects, undistinguishable.

t Returned in generic terms, as Abyssinian, Negro etc.

II.

Religion

Population

Proportion
to 100,000

I. Religions native to India . . .

A. Tribal Animism

B. Offshoots of Brahmanism.

(i) Hinduism

(2) Brahma and Arya Samaj . .

(3) Sikhism

(4) Jainism

(5) Buddhism

II. Religions of Foreign Origin . .

C. Mazdaism

218,797,808

8,176,560

206,715,341

96,054

2,185,330

1,333,820

290,703

64,193.255

93.449

14,436

61,315,475

2,767,235

2,660

77,316

2,890

73,046

34

772

471

103

22,684
33

D. Judaism

5
21 667

E. Islam

F. Christianity

G. Others

97S
I

Total . .

282,991,063

100,000

§ 12. One of the most interesting ethnographical questions entering
into the Census inquirj' is that of the rate at which Brahmanism is in
name, at least, absorbing the Animistic tribal population. Unfortunately,
this cannot be fully solved from the returns, owing to the different inter-

Social Organisation. A. Historical.

pretations given to the instructions for recording tribal creeds and languages.
The enumerators, or those who instructed them, adopted somewhat ar-
bitrary standards of orthodoxy and philology, and what was set down as
tribal in one tract appeared under the more general title in another, just
across a political frontier. Speaking generally, the tendency seems to have
been to return the tribal terms wherever the community in question is
in predominant occupation of a cuntinuous and well-defined region, and is
thus in comparative isolation from the civilisation of the plains. Where,
on the other hand, the tribe is interlaced with the Brahmanical peasantry,
the distinction was less noticed, and probably the line is in reality less
discernible. It may be interesting, in spite of the above drawbacks, to
learn what the conditions are as set forth at the Census, so a further
Table, in which the proportion in which each tribe returned the tribal
religion and language is given in the Appendix.

SOCIAL ORGANISATION.

A. Historical.

§ 13. Tribe. — In the outline given in the Introduction it was shown
that throughout the greater part of continental India, the region most in-
fluenced by foreign blood, distinctions of race have been practically effaced
by centuries of cross-breeding. It is to be noted, however, that wherever
a race can still be geographically demarcated from its hybrid neighbours
the ethnic constitution tends to be tribal, consisting, that is, of groups
with a common name, the tradition of kinship or descent from a common
ancestor, human, demi-god or wild animal, as the case may be, and
claiming or occupying a definite territory. The system on which the tribe
is organised varies considerably according to the race and the conditions
under which it lives. That most intimately connected with India proper
is found amongst the Kol-Dravidians of the Central Belt. Here, the tribe
is subdivided into numerous exogamous sections, each bearing the name
of a plant or animal of the locality, and marrying almost invariably within
the tribe itself, or, at most, not beyond an adjacent and probably kindred
community of similar organisation and form of religious and domestic ce-
remonial. The Mongoloid tribes of Assam and the eastern frontier are also
divided into sections professing blood -relationship, and therefore not
marrying within the section, but trusting to their fellow-tribesmen of other
divisions to provide them with brides, either by arrangement or capture.
On the opposite frontier, the tribal constitution of the Pathan and BalOch
races is of a markedly different type. The Baluch tribe is bound together by
political rather than ethnic ties, owning allegiance, that is, to a common
Chieftain ; but amongst the clans which go to form this unit, there is found
very often, if not usually, the tradition of blood-kinship, surrounded by a fringe
of strangers who have affiliated themselves to the community for the purpose
of mutual defence, and who, after a term of probation, are admitted to full
tribesmanship. The subdivisions of these clans are exogamous, and there
is a tendency, but nothing stronger, towards endogamy within the tribe.
Amongst the Pathans the tribe is more closely knit, and the bond is
kinship in the male line. As amongst the Baliich, however, strangers are
admitted to qualified membership, tending, in time, to be treated, by fiction,
as kinship. There is not the element of allegiance to a common Chief,

10 5- Ethnography.

though in many cases such dignitaries do exist and are regarded as war-
lords and representatives of the tribe in dealing with the outer world.
But the internal management uf tribal affairs is vested in a tribal Council,
composed of the Heads of clans or other subdivisions of the main body.
Marriage takes place, as a rule, within the race, and in practice is re-
gulated by Muslim, not tribal, prescriptions regarding affinity. The in-
fluence of these races, especially of the Pathan, upon the whole population
of the western Panjab, has had the result of substantially modifying the
social structure, elevating the tribal, or blood connection, enlarging the
marriage field, and generally promoting the adoption of the freer life of
the Highlands in preference to the stricter and more elaborate system
which prevails throughout Brahmanic India.

§ 14. Caste. It is with the latter, however, that this review is mainly
concerned, and the only object of the above remarks is to differentiate
the organisation of, so to speak, the pure races of India from that of the
great mass of the population. Amid the bewildering variety of the com-
plicated civilisation of this last the one and only characteristic which can
be said to be universal is the sentiment which underlies the scheme of
life upon which the whole of the social edifice is based and its component
parts are respectively distinguished and coordinated. This sentiment,
moreover, may be said to be the very spinal cord of the main religion
of the country, supplying the vitality and support which neither doctrine
nor ritual are sufficiently coherent to provide. By its means, Brahmanism
has become, as has been said by a competent observer, "a way of life,
"interwoven into the whole of existence and society; placing every na-
"tural habit and duty upon a religious basis so entirely that it is impossible
"for a Brahmanist to draw a distinction between sacred and profane. A
"man's religion means his customary rule of every-day life. His whole
"social identity belongs to his religion". (Lyall. Asiatic Studies.) This
omnipresence of the religious sanction and the rigidity which it imparts
to diversity elsewhere susceptible of diminution or effacement is not
only the most prominent feature of the social organisation of India,
but is also peculiar to the latter, marking it out as distinct from any
other civilisation in the world. In other respects, there is little in the
system which is not to be found, or which has not at some time or other
existed, in other countries, even of the West, though it has there been
long ago worn away by other influences. The crystallisation of certain
bodies into definite orders or classes, for instance, is a common, almost
a universal, trait, and amongst them the tendency to become hereditary
and as exclusive or aspiring as circumstances allow may almost be called
natural. A superior and conquering race, again, has been known elsewhere
to settle for generations alongside of a population in every way inferior
to it, compelling the latter into servile conditions and drawing upon it for
wives and concubines without making any return in kind. Sacerdotalism,
too, has had its day of supremacy elsewhere than in India. Restrictions
in regard to the choice of a wife and upon participation in meals of a
commemorative or other ritualistic significance, are, of course, common
property. But in no other case has the position of a sacerdotal class been
so firmly established nor has its influence so deeply permeated the whole of a
vast community, as to enable it to prescribe, under the sanction of religion,
a code of elaborate prescriptions on domestic and personal conduct which
is accepted by all as the ideal, according to the relative conformity with

Social Organisation. A. Historical.

which the rank of every group of the society, from top to bottom, is
unalterably settled. A system of this description, which, practically un-
changed in its main principles, has for many centuries regulated the lives
of millions; which is absorbing every generation more and more of the
tribal population of a lower type brought into contact with it, and which
has not only successfully resisted, but has even been to a great extent
assimilated by so dogmatic and uncompromising a rival as Islam, must
obviously have its roots very deep indeed in the proclivities and traditions
of the multitudes living under it.

Whether it be indigenous to India, or whether it existed in an em-
bryonic form amongst the Aryas before their great dispersal, is a question
which has been the subject of wide and erudite discussion. Probably it
is insoluble, most theories of primitive society being apt, according to
Sir Henry IMaine, to land the adventurer in a region of mud-banks and
fog. This, remarks the author of the last Census report (1901, p. 546),
"is more especially the case in India, where the palaeological data available
"in Europe hardly exist at all, while the historical value of the literary
"evidence is impaired by the uncertainty of its dates, by the sacerdotal
"predilections of its authors, by their passion for wire-drawn distinctions
"and symmetrical classifications, and by their manifest inability to draw
"any clear line between fact and fancy, between things as they are and
"things as they might be, or as a Brahman would desire them to be".

§ 15. The social divisions which form the units of the system in
question are known in the West by the name of Castes, which was given
them by the early Portuguese travellers. It is said to be derived from
the Latin word casta^ pure or unmixed, in itself connoting segregation,
and was applied by Camoens, for instance, in the sense of tribe or even
race, to the Pulayan or helots, in contradistinction to the Nayar, their
conquerors. It needs but a very short time in the country to bring home
to the most casual observer the ubiquity of the institution, and to make
him acquainted with some of its principal exoteric features. He might
possibly feel himself in a position to define it, an enterprise from which
after longer experience he would shrink, as the more caste is studied,
the more numerous are the qualifications found to be advisable in describing
it. It is necessary, however, for the purposes of this review, to set forth
in terms as definite as the case allows the leading features of the com-
munity which forms the main subject of this work. Of the many definitions
which have been given by various authors, the most satisfactory, on the
whole, is that adopted by Mr Gait, the joint author of the last (1901)
Census Report, in dealing with the castes of the Province of Bengal,
"A caste", he says (p. 354), "is an endogamous group or a collection of
"endogamous groups, bearing a common name, the members of which
"by reason of similarity of traditional occupation and reputed origin are

"generally regarded as forming a single homogeneous community,

"the constituent parts of which are more nearly related to each other
"than they are to any other section of the society". From this it appears,
then, that the members of a caste may only marry within its limits; but
nearly every caste is made up of sections upon whom the same restriction
is imposed with reference to their limits, the title of the subdivision being
added to that of the main aggregate. The occupation, again, which is
common to the latter, is a traditional one, and is not by any means neces-
sarily that by which all, or even most, of the group make their living

12 5. Ethnography.

in the present day. On the other hand, the common origin, which is now
claimed by most, is largely a matter of fiction, accepted, however, without
cavil. The factor of public opinion, too, is of some importance in the
definition, since the view taken by an aspiring section of a caste of its
relationship to the main body is apt to differ from that accorded to it by
the other castes amongst whom its lot is thrown, whilst the acquaintance
of the upper classes with the organisation of those below them, and their
interest in it are of the slightest, until perhaps an encroachment comes within
measurable reach of their own position. It sometimes happens, therefore,
that a subdivision by retaining its own title but substituting a fresh one
for that of its main caste, obtains a jumping-ground for a new start in
society, which may impose upon the outer world but not upon the imme-
diate surroundings. Reverting, for a moment to the definition, it may be
noted that while endogamy is the chief characteristic of the organisation,
an exception is found in the case of the Rajput, or military caste, which
is based upon exogamous clans or tribes. These have in many cases fixed
their own circle of intermarriage within the caste on considerations other
than those current amongst the rest of the Brahmanic community. There
are apparently ethnic reasons for this peculiarity, to which reference will
be found below.

§ i6. The caste system being an institution essentially and exclu-
sively Indian, the question arises whether its origin is to be sought amongst
the Aryan immigrants or to be ascribed to those whom they found in
possession of the field. Or, again, assuming that it is the resultant of the
contact of the two social systems, what is the influence respectively attri-
butable to each? The view now very generally held is that it is the
product of no single cause, but that to its establishment in the form in
which it now prevails, several factors, Aryan, pre-Aryan and hybrid, have
at different times contributed. Of these by far the most prominent is the
hieratic influence by which the main principles of the system were fixed
and the standard set by which social position is graduated. That influence
derives its authority entirely from the Vedic tradition, so it becomes neces-
sary to see what information is obtainable from that source regarding the
social organisation of the community amongst whom it originated. As in
regard to all else concerning the earlier life of that community, reference
must here be restricted to the Suktas of the Rksarhhita. These composi-
tions must of course be defective in some respects, and from their character
and the occasions they were intended to serve they cannot be expected
to furnish a complete and detailed picture of the organisation of the body
to which they relate. Nevertheless, the general conditions of life among
those peoples were simple, and the relations between those who offered
the sacrifice and the divine power whose good offices were solicited
through it were so intimate and practical, that from the large collection
of effusions handed down to posterity a very fair general notion can be
formed of the leading facts relevant to the subject under consideration.
§ 17. It appears, then, that at the comparatively advanced stage of
progress which the Vedic Aryas had attained by the time represented in
even the earliest invocations of the collection, the community was or-
ganised into clans, or groups of related families which, in turn, were
collected into tribes, to which the clan was subordinate. Various other
terms are met with imi)lying subdivision of either tribe or clan. They all
refer to a pastoral life and indicate a by no means high degree of cohesion.

Social Organisation. A. Historical. 13

Alongside of these sections were two classes or orders, evidently of later
development: the nobles, headed by a Chieftain, and the ministers of
religion, who conducted the public sacrifices. The mass of the community
below these orders is collectively referred to as the "clans", or "peoples",
always in the plural. The Family, as a unit, was strongly developed. Its
worship was purely individual, strictly secluded from that of its neighbour,
and conducted in private by the Paterfamilias conjointly with his wife.
The tribal sacrifices were open to the "clans", and were conducted, at
least in the stage to which the Suktas relate, in the presence of the
Chief of the tribe, by a priest acting on his behalf. It seems probable that
the ritual had by then reached a pitch of complication which necessitated
the employment of trained professionals, but the performance of this act
of faith was not otherwise the exclusive privilege of the sacerdotal class,
for occasionally scions of ruling families officiated, and there are cases
in which the right of the priest was disputed by others. It is obvious,
however, that the duties fell more and more into the hands of trained
experts, irrespective of the personal separatism which tends to attach itself
to a sacrificial priesthood, as the ceremonial became more elaborate, and
still more, after the invocations which accompanied it had ceased to be
improvised and the compositions of the older Psalmists were recited in
a regular liturgy. The experts closed their ranks against the layman, and
became a class by themselves, whether they maintained their numbers by
heredity or recruitment. It may reasonably be assumed, too, that the order of
nobles, especially in the case of tribal chieftains, would gradually tend
towards a hereditary character, though the frequency of intertribal strife
and the migratory life of the communities militated against the con-
solidation of political authority in such hands.

§ 18. So far, it may be observed, there is nothing in the above more
or less hypothetical social organisation of this branch of the people con-
ventionally called Aryan which materially differs from what is known to
have prevailed amongst the others branches of whom the early history
is on record. It was after the Vedic tribes had debouched upon the plains
of north-western India that their social system assumed its unique and
special features. Here, two new factors awaited them, each being insufficient
by itself to determine the future course of their civilisation, though the
combination of the two led to that result. The immigrants came into
contact, in the first place, with a race far below them in physical and
social characteristics; and they found themselves, in the second, in the
presence of a vast and fertile expanse of country over which the inferiority
of their opponents allowed them to spread freely. Whatever may have
been the difficulties in dealing with the Dasyus which were at first ex-
perienced by the Aryas, the superiority of the latter ultimately asserted
itself in an incontestable manner, and those who resisted them were either
reduced to subjection on their native soil, or rolled back before the ad-
vance of the new-comers. That the Aryas failed to take advantage of
their opportunities to establish themselves upon a national basis appears
to be ascribable to the fact that, except in race, they were any thing but
a homogeneous body. Tribe was constantly at war with tribe, and in their
slow onward progress there had been no signs of combined general effort.
It is true that after they had been some time in the plains larger aggre-
gates were occasionally formed by military Chiefs, but they were unstable
and perpetually being dispersed and re-formed in the vicissitudes of tribal

14 5- Ethnography.

contests. The stable element, then, in the colonisation, was not supplied
by the Court and its army, but by the village. This community seems to
have been an institution of very early date amongst the Vedic tribes, and
was established upon a clan, or even a family, basis, cemented by the
possession of a definite tract of pasture or arable lapd. The opportunity
for forming detached and independent settlements of this kind was fa-
vourable. Land was plentiful, and whilst the supply of menial labour was
provided by the Dasyus retained in subjection upon the soil of which they
had been dispossessed, the danger of reprisal by the rest was removed as
the more adventurous bodies of the Arj-as extended their frontier further
and further into the interior. The necessity of combination for mutual
defence against the alien waned therefore into insignificance. The tie of
tribe, never very strong or well defined, would naturally be subordinated
to that of territorial ownership, especially if the smaller unit were founded
on blood-relationship and settled communal interests, and there was no
common end which made an urgent appeal for collective action. In these
circumstances, the dispersal of the original Vedic communities far and
wide under new and more prosperous economic conditions tended towards
the development of a parochial separatism, which possibly the presence
of large bodies of alien helots may have helped to divert from wider
political conceptions. The village community being left, on this hypothesis,
to itself, organised its members on lines suggested by its requirements,
which multiplied, of course, in proportion to the increased resources af-
forded by a settled life. At the head of the social scale stood, as now,
the possessor of land and beeves; at the foot, the stunted and swarthy
alien. Between these extremes room had to be found for the increasing
number of handicraftsmen, as well as for the hybrid progeny of the Arj-a
by Dasyu women. What with the absorbing interests of this bucolic
microcosm, and the absence of any specially powerful motive for political
combination into larger units, the gap between the masses and the military
dominant class tended to widen, and the fortunes of the ruling houses
became a matter of comparatively little importance to the village. There
remained, however, the tie of race. Whatever may have been the strength
of this in pre-Vedic times, it became very prominent, as has been stated
in the Introduction, when the Aryas came into collision with the Dasyus.
The one term used collectively of the whole of the former community is
the "colour" of the Arya as contrasted with that of their foes. In the in-
vocations, until, that is, a period is reached when bodies of other and
non-Vcdic Aryas appeared upon the scene, this characteristic is made
practically equivalent to worship. The worship, in turn, was that of the
Family, originally expanded on special occasions to the sacrifice offered
under the auspices of the Chieftain for his tribe. The latter ceremony may
easily have waned without affecting the essential daily rites of the house-
hold, to which, indeed, the dispersal of the tribe and the constant presence
of the Dasyu helots at the gate might be assumed to lend additional value.
Nor, again, would the expansion and re-formations of the Ar^-an community
tend to diminish the influence of the professional, or Brahmanic, ministry.
This had probably grown into a closed body before the dispersal, but it
was attached in the first instance to the person of the Chieftain, and
obviously could not be otherwise than dependent upon those on whose
behalf the priestly offices were undertaken. The Brahman, then, was bound
to follow the fortunes of the rest of the community, and scatter as they

Social Organisation. A. Historical.

did. They, in turn, could not well dispense with the services he alone
was competent to render. The language of the invocations had become
obsolete, but texts from them were an essential part of everj- ceremony,
and had passed, it would seem, into the stage of spells, potent only in
the mouths of those who had professionally learnt them, a class which
had taken care to prevent others from particijiating in that advantage.
The value of this qualification increased, naturally, as the various bodies
of those who placed their faith in it receded further from their traditional
race-unity. There were other conditions, too, favourable to the growth of
sacerdotal influence, and to the transfer of the attention of the hieratic
order from the fluctuating fortunes of the military aristocracy, (by whom,
moreover, its exclusive and privileged character was by no means uncon-
tested,') to the more amenable medium of the incoherent democracy of
the village, where the circumstances were evidently open to organisation.
A good foothold was provided in the high value placed upon the
purity of the family blood, the maintenance of which was the predominant
object of the Vedic social system, as it seems to have been that of other
Arj^an communities in their early days. The ideals and practice of the
upper classes in regard to such a question constitute the hall-mark, as it
were, of gentility — in the older sense of that term. Their natural ten-
dency, accordingly, is to filter downwards through the society, each section
adopting, as it attains a secured position, some measure of precaution
against degradation through admixture with bodies which it considers its
inferiors. Whether this sentiment of exclusiveness hardens into separatism
or is merged in wider conceptions depends upon the circumstances in
which the community happens to find itself during the early period of its
settled existence. Pressure from outside may necessitate a political orga-
nisation which reacts upon the domestic structure, or the struggle for life
within the community itself may tend towards a more comprehensive
grouping. In the advance of the Aryas into India neither of these motives
seems to have been predominant. The way was open, therefore, for the
confluence of the two peaceful currents which had throughout all vicis-
situdes preserved their continuity — the sentiment of family purity and the
hieratic administration of the ancestral worship. In regard to the former,
the foundations of a closed order based on heredity had been laid, as
mentioned above, amongst the priests and the nobles, at a verj' early
period, and the bias in favour of such distinctions amongst the "clans" was
necessarily accentuated by the contiguity of the dark races, on the one
side, and the evolution within their own community of occupations un-
recognised, because unknown, in Vedic tradition. Manual industries, it
should be borne in mind, were invariably depreciated by the Arya of the
west, where they were relegated to the servile population ; and in India,
whether they were carried on by the Dasyu, the half-breeds, or the poorer
members of the Clan, they could not fail to bring into prominence the
possibility of contamination or abasement of position, either on racial
grounds or by reason of the inherent or conventional impurity of the
calling. In these circumstances, the idea which seems to have been adopted
to prevent the flowing tide of impurity from submerging the cherished
landmarks of pride of family and of race, was to establish an alliance
between conventional purity of race or calling with the ancestral religion
of which the Brahman was the sole exponent. The situation could be
stereotyped by the establishment of the distribution of society upon divine

i6 5. Ethnography.

ordinance. It is true that as is now generally admitted, Caste, still less the
Caste-system — which is the subject now in hand — did not exist amongst
the Aryas of the Sukta period. The materials for it, however, had been
provided by their descendants, and it only remained for the Brahmans,
who were now in a position of power in the interior, to set their seal
upon what they found ready to hand. The Purusa-Siikta of the Rgveda,
decreed by modern scholars to be the product of the latest Vedic period,
verging upon that of the early Brahmanic supremacy, is the Magna Charta
of the caste system. In this composition, a divine origin is ascribed to
four classes, the social position of each of which is thus irrevocably fi.xed.
The two first are the Vedic orders above mentioned. Then comes a third,
the title of which is derived from the Vedic term for the "clans" in the
aggregate, whilst a place of degradation is made for the lower orders
generally, in which, apparently, though the point is not certain, is merged
the Dasyu community. Into this strictly demarcated classification were
compressed all the numerous sections of the population existing at the
time when the Brahman Procrustes undertook its application to the facts
of everyday life. In such an arrangement it is obvious that the leading
place in the social hierarchy would be assigned to the Brahman, and that
any encroachment upon that supremacy would be amply provided against
by the establishment of the principle of heredity in determining rank.
Endogamy is here implied, as it is essential to the preservation of the
family or caste purity that the mother of the heir should not be the medium
by which any taint can be introduced into the blood. The principle under-
lying the scheme of organisation seems to have received universal recog-
nition, possibly because the standard of purity in regard to function had
already been fixed by public opinion, whilst that applied to social inter-
course, being bound up to a great extent with religious ceremonial, would
be graduated in accordance with the example set by the class which
prescribed or regulated that branch of caste duty. It seems doubtful, indeed,
whether the two lower classes of the Brahmanic scheme ever had more
than a literary existence, and were not a convenient expedient for severing
the masses from the privileged classes. As a further security against a
rivalry which in after times, perhaps through Buddhism, became trouble-
some, the Brahmans, in due course, proclaimed the Ksatriya order also
to be extinct.

§ 19. Assuming the above hypothesis to be well founded, it is clear
that whilst the system upon which Indian society is organised is due to
the influence of a hereditary priesthood, which acquired thereby a position
of unparalleled supremacy, there is no need to "smell Jesuitry" in the
history of its genesis, and to brand it as nothing more than the full-blown
device of subtle and self-regarding Brahmanism. It appears, in fact, that
the sacerdotal element in its elaboration was met at least half-way by
the inclinations of the lay public, as evinced by the form their civilisation
had begun to assume. The sacrosanct position of the Brahman being
once established as the pivot of the system, the development of the latter
precceded on the lines indicated by the code of purity adopted by the
priestly order. Recognition of the inherent sacredness and spiritual autho-
rity of the Brahman became essential, and even the great sectarian move-
ments in derogation of the exclusive privileges of the sacerdotal class
left caste untouched, and ended, accordingly, in the actual, if not nominal,
acceptance of that condition as the inevitable apex of the system they

Social Organisation. A. Historical. 17

retained. Doctrinal orthodoxy, indeed, could not have had much weight
in the social balance after the pantheon had been enlarged to admit the
claims of popular local deities, and the non-Aryan beliefs and ritual had
been adapted to the flexible requirements of the Brahmans. Schism on
religious grounds occurred, no doubt, in the earlier times, as it has con-
tinued to do, and fresh subdivisions were formed in consequence, but
these involved no change in caste or social position unless they happened
to entail the violation of prescriptions relating to the purity of the family
or the individual. These prescriptions are the operative part of the system,
regulating as they do, marriage, food, occupation, and intercourse with
the rest of the community. They are thus of a quasi-public character and
the breach of any of them brings the stigma of pollution not only upon the
individual but upon the family and the castefellows who come into contact
with the offender. They are, moreover, comparatively easy of detection,
and are thus well within reach of the discipline of the caste tribunal, a
consideration of some moment amongst the masses, with whom detail
looms higher than in classes where tradition is stronger and position
more assured.

Other factors contributed, of course, to the consolidation of the
system; and amongst them have been included the devout belief in the
omnipresence of supernatural agency, permeating all classes from top to
bottom, and predisposing them to submission to priestly authority. There
is, again, the doctrine of metempsychosis, which, regarding the present
as the direct heir of the past, lends valuable support to the notion of
predestined lot in the successive births into this world to which all are
subject. The apathetic character of the people, also, is taken into account,
and the marked absence amongst them of the "noble discontent" with
their circumstances which spurs men on to efforts to improve them. It is
open to question however, whether the two last preceded the institution
of the caste system, or not. Be this as it may, the most potent factor is
the Brahmanic standard of purity, and the desire to emulate it. This is
the thread upon which is strung the astounding collection of otherwise
independent communities into which the population of India is now divi-
ded and which multiplies almost every year the number of its units. It
underlies the demarcation by race, in the form in which it chiefly prevails,
whether amongst the village menials or the Hill tribes which have become
or are becoming, castes, as they abandon customs which are incompatible
with it. In function, again, which in its lower grades is closely connected
with race, the social gradations are based upon the relative cleanliness
of the pursuit, though not without a glance at the ancestry of those
who have taken to it. Religious differences only lead to the formation
of a separate caste, when as above indicated, they are accompanied by
a departure from the social observances of the original body, upwards,
it may be, or in the opposite direction. The constant multiplication of
castes, indeed, is attributable for the most part either to the assumption
by a section of an existing caste of a higher standard of purity than the
rest in occupation, marriage regulations, or food; or, on the other hand,
to the excommunication of a section from fire and water for a violation
of the caste rules regarding such matters. This fissiparity of castes is a
subject of great intricacy to which space does not allow more than cur-
sory reference here. It is necessary to make some mention of it, however,
in order to show that, rigid and compressive as may be the framework

Indo-Aryan Research. H. 5. 2

iS 5- Ethnography.

of society imposed by the caste system, it does not preclude mobility
within the multitudinous cells of which it is composed, and provides, too,
for the increase of their number by accretion from outside. It is perhaps
still more important to note that the converse process does not take place.
A section once split off does not rejoin, nor do different castes coalesce
with each other to form larger communities of the same character. With
the object of illustrating these features of the system in actual operation,
a brief description of the more representative castes has been included
in the latter portion of this review. Through this more information may
be gained, it is hoped, than can be conveyed by a series of general
statements, each of which, like most general statements concerning India,
requires abundant qualification to meet local exceptions. It must never
be forgotten that India is not a country but a collection of countries, and
though caste as an institution is universal, and the basis of the system
which has been the subject of the foregoing review is the same throughout,
the form assumed by the superstructure raised upon that foundation differs
materially in different regions. If any generalisation be sustainable, it
would be, perhaps, that caste tends to be strong where the population
is generally prosperous, and also where the system was adopted after it
had reached maturity among those who were the means of introducing
it. It tends to be weak, on the other hand, where the means of sub-
sistence are less abundant, and occupations, therefore, cannot be so
strictly demarcated as they are under more favourable conditions. The
stage of civilisation, too, attained by the time Brahmanisation set in, seems
to have been a factor of some weight in determining the extent to which
recognition should be accorded to local customs and beliefs.

t; 20. Thus, in the south-Dravidian part of the peninsula, the caste
system flourishes in full vigour; but it has simply been engrafted upon
Tamil institutions, and, as far as the masses of the people are concerned,
little change has been effected by it in their food or their special regu-
lations regarding marriage ; still less in their worship, in which the Brah-
man takes no part except where one of the more powerful of the local
maleficent goddesses has been adopted as a manifestation of some Puranic
divinity. The lower orders there occupy a position of degradation differing
from that of the corresponding castes further north in that a good many of
them do not accept it; and having a working tradition of former power, if
not supremacy, they are continually making efforts to get their claim to a
higher rank recognised by their actual superiors. The subdivisions among
them increase accordingly. On the other hand, the artisan castes are here
found united to an extent unknown in the present day elsewhere. This
combination is of long standing, and is probably the origin of the Right
and Left-handed distribution of castes which is only found amongst the
Tamil people. The South, again, having always been fertile in sectarian
disputes, doctrinal schism amongst the local Brahmans has resulted in
some instances in separation in social intercourse, another development
not found elsewhere. The Rrahmanism of Tclingana has considerably less
of the pre-Aryan clement, left in it, probably because there was partial
colonisation of the Andhra region through Orissa or otherwise, by immi-
grants from the Ganges valley, before the Dravida region was reached.
The inhabitants, accordingly, though lax in their observances compared
to the Brahmanists of the North, consider themselves higher in position
than the Tamil castes, and when settled amongst the latter, avoid inter-

Social Organisation. A. Historical. ig

mixture as far as possible. The greater prosperity of the South, however,
has given to its caste system a strength and complexity not found in the
present day in the less favourable conditions of the upland tracts. Along
the East coast the Tamil features prevail almost till they join the Orissa
system, which, probably from the isolation and the timid character of the
population, has the reputation of being the most bigotted and priest-ridden
of its kind. In Lower Bengal, the system is an exotic, as in Madras, and
was introduced long after it had reached maturity in upper India. It took
root however, under different auspices. The country was occupied by the
Aryas or their hybrid descendants in the course of their general expansion
down the valley, and the population encountered consisted of the wild
tribes of the forest or amphibious dwellers in the Delta, Kol or Mongoloid,
easily subjected, like the Dasyu of the north, and not, like the Tamil
communities, long settled on an agricultural basis, to be approached by
missionary enterprise only, not by armed force. The subject classes seem
to have been left to assimilate their organisation to that of their superiors
without tradition or authority to guide them. When, at length, the official
graduation of society was taken in hand by one of the more powerful
local rulers, the flood of Islam overran the country before the new re-
gulations had time to gain foothold amongst the people. It appears, there-
fore, from physical features and the titles of caste subdivisions that bodies
were formed either by race, afterwards split up by function, or by com-
munity of function overriding race differences and often determined by
locality. The relations between these bodies, therefore, are more than
usually indefinite, and owing to the absence of a landholding aristocracy
of the military order and the comparative weakness of the Brahman immi-
grants, changes or claims to change of rank are more frequent here
than in any other part of India. Amongst the lower classes these pre-
tensions are usually based, as in the Tamil country, upon tradition, often
not without foundation, of a former position far above that now assigned
to them. For generations they have been deposed, but the prosperity they
enjoy in modern times induces them to revive their dormant claim. Still
more immature in its development is the caste system, if so it may be
called, which prevails in the Assam valley. Setting on one side modern
immigrants from Bengal and the Brahman, there is but one community of
even nominal Aryan origin. It is now held to represent the early Aryan
immigrants, who reached the seclusion of the Brahmaputra valley before
the caste system had been developed in Bengal or wherever these co-
lonists originated. In their case the development was apparently retarded,
first, by pressure of Mongoloid tribes around them, conducing to a united
front; and, later, to the hold which Buddhism obtained for some time
over this tract. The above caste, or racial community, included all the
ordinary professions but they were not formed into castes, and even
now that process is by no means complete. Even the higher classes are
lax, too, as to intermarriage, and visit the mesalliance of a girl upon her
individually, not upon her relations, as would be done in other parts of
India. The Brahman, too, falls into line with the rest, and disregards the
stricter rules of his order as to marriage. Special arrangements exist for
the incorporation into castes of the indigenous population; and the fa-
cilities they afford for a subsequent rise in rank on increased observance
of conventional purity are unwontedly liberal. The same spirit is manifested
in the relations between orthodox Brahmanism and the Kol and Dravidian

20 5 Ethnography.

tribes of the Central Belt. The tribes of Chutia Nagpur tend to get merged
into the Bengal system, and those of the Satpura and Vindhya, where
conversion seems to lead to more comi)letc breach with the older regime,
gradually mix with the lower castes of cultivators in the ])lains. Between
the Jamna and the Ghogra or even the Kosl, the caste system seems to
have developed upon what may be termed more normal lines than in any
other part of India, as is, perhaps to be expected from the proximity to
its birth-place. The process of evolution was seriously interrupted, however,
by the Muslim occupation, which scattered the leaders of society and
swept away many old landmarks. In course of time, the old order was
reestablished in full force, though the traces of the cataclysm have never
been (juite effaced, esjjccially amongst the functional castes. It is worth
noting that in the u])per Jamna tract and well into the eastern Panjab
caste remains entirely unaffected by conversion to Islam. It is held by
some, indeed, that by the elimination of the Rajput, or fighting man, the
Muslim left the way more open to the Brahman, whom they disdainfully
ignored. At all events, the present social conditions of the region longest
and most absolutely held by the IMoghal regime appear to confirm con-
clusively the evidence afforded by the relations between Brahmanism and
the pre -Aryan worship of the south and centre, to the effect that the
hold of caste upon the popular mind is altogether detachable from reli-
gious doctrine, and rests, as indicated above, upon its social restrictions.
In the western Panjab caste is weaker than in any other tract, and this
seems to be attributable to the combination of two influences. First, there
is the tribal sentiment, derived from the vicinity of the Pathan and Baluch,
referred to earlier in this work. It found a ready acceptance amongst
the Rajput and Jat races of the plains, who were themselves organised
upon a tribal basis, with a lightly worn veil of caste thrown over the
arrangement. Then, again, the struggle for life in a comparatively infertile
country conduced to the mobility of occupation to an extent seldom ne-
cessary in the richer tracts to the eastwards. The adoption of a lower
class of calling under pressure of need leads, of course, to the loss of
social position, but not, as it would on the Jamna, to excommunication.
Caste is also weak in the lower Himalaya, but for a totally different reason.
These valleys are the only tracts to which the Muslim never penetrated,
and, under the auspices of refugee Rajputs, society is there constituted
upon a system untouched by foreign influence. The Chief is emphatically
the fountain of honour, and can uplift or degrade a caste or even a family
as he pleases. In the Panjab Hills, therefore, caste is remarkably fluid.
Every community above the menial aspires to rise by some means or
other to the rank of that above it, whilst it takes wives from and eats
with, that immediately below it.

The various tracts which have been mentioned present the most
strongly marked peculiarities in their caste systems, but in each of the
rest there will be found certain characteristics in which it differs from
others. Into these it is not proposed to enter except cursorily. In Sindh,
for instance, the whole population embraced Islam, and the only large
indigenous Brahmanic caste left is that of the traders. The rest, however,
have maintained both racial and functional divisions regulated generally
on caste lines. The adjacent peninsulas of Gujarat have been frequently
occupied by aliens, and this fact, together with the fertility of the main-
land, tends first, to great subdivision of castes, the titles of the sec-

Social Organisation. B. Descriptive.

tions indicating intermi.xturc of races as in Lower Bengal, and then to
strict observance of caste discipline, as in the Gangetic region. The Konkan,
too, has had from time to time a strong influx of foreign Brahmans, and
this, along with its isolation, have helped to rivet firmly the priestly yoke
upon the people. In Rajputana, too, as is natural considering the history
and character of the ruling classes, Brahmanism is in high honour, though
the difficulty of making a living in the desert portion of the tract allows
a latitude of occupation among the poorer castes similar to that which,
for the same reason, prevails amongst the probably kindred tribes of the
middle Indus.

Distinctions such as these are illustrated as far as space allows in
the following pages of this work, where, in the description of its main
constituent parts, is shown in actual operation the system of which the
development and conjectural origin have been outlined above.

B. Descriptive.

§ 21. Regarding the subject in its ethnographic aspect, it is obvious
that it must be a task of extraordinary, almost insuperable, difficulty to
reduce to anything like accurate numerical terms the component parts of
so vast and complex an organisation as that sketched above. It should be
borne in mind that the object of the Census is to obtain a record not
only of scientific value in the service of ethnography, but of practical
importance in the every -day administration of the country. The social
position and the numerical strength of different sections of the community
are essential facts in connection, for instance, with public instruction or
with measures for the promotion of the comfort or convenience of the
locality. The Courts of Justice, again, are frequently called upon to decide
questions of rank or privilege in which the relative numbers of the litigant
parties are points relevant to the inquir>-, and which cannot be safely left
to the evidence of the disputants, in view of the "megalomania" which is
probably at the bottom of the whole controversy. Even the identification
of an individual cannot be satisfactorily established in the case of many
of the more important social divisions by less than two or even three,
successive questions, and often the credibility of a witness is decided by
a casual detail of caste convention. On these considerations, and with an
eye to the known probability of error in the direction of either excessive
generality or excessive minuteness of description, provision was made at
the Census for the return of social divisions under two headings, first, the
main body, such as caste or tribe, and, secondly, the subdivision to which
the individual may belong. In the larger communities, indeed the latter
is the more distinctive designation, and was adopted, accordingly, as the
unit of compilation in the returns prepared for local use. Lower than this
it is unnecessary', for administrative purposes, that the inquiry should go;
but it must be recognised that from the ethnological standpoint, the more
minute subdivisions of the community are often more pregnant of suggestion
or information than those of which they form a part, and must be adequately
dealt with in any special investigation, such as that now engaging the
attention of those employed upon the Indian Ethnographical Survey.

It must also be understood that neither the Provincial nor the Imperial
returns claim to present anything beyond a partial and very imperfect
picture of the astounding fissiparity of the Brahmanic social system in the

22 5 Ethnograph\-.

full vigour of its present existence. The Imperial Table, even after a
somewhat drastic process of compilation, contains nearly 2,400 separate
items, and the project of expanding it to the full limits of the subject
inevitably calls to the memory of the expert the concluding verse of the
Gospel according to St John. Take, for instance, the feature of endogamy
alone. Every subdivision recorded in a Provincial Table, covers, if the
main body be widely spread, many others, none of which intermarries
with the rest. Not only so, but the main body itself does not recognise
any social tie with the body bearing the same name located in a distant
part of the country, even though, as sometimes, happens, the same verna-
cular language may be spoken by both. Each of these local subdivisions,
moreover, is divided into its respective endogamous sections; some of
them professing a different religion, and occupying, perhaps, quite a
different position in the social hierarchy of the neighbourhood from that
of the synonymous section elsewhere. Even the Provincial groups, therefore,
subjoined to the general aggregate in the Table, convey an impression of
homogeneity not in correspondence with the actual fact.

§ 22. With the above qualifications and reserve, then, the figures to
be found in the Imperial returns must be taken as providing as trustworthy
information as is now available upon this branch of the subject. In the
Tables, the items are arranged in alphabetical order, a form of record which
has its advantages from an official point of view, in that it raises no
awkward questions as to position or precedence; and, if accompanied, as
in the Madras list, by a brief practical account of the principal divisions, it
is useful for reference on individual cases. There, however, its function ends,
and some form of coordination becomes necessary before all these isolated
nuggets of information can be got to collectively yield their tribute towards
the common object of illustrating the main characteristics of the social
organisation of the different regions of India. It is as well to admit at the
outset that in view of the varied origin and history of the social divisions
in question and of the various forms the social system has assumed, no
classification upon a single a definite principle is possible. It is equally
judicious to assume that, taking into consideration the diverse and often
mutually inconsistent theories held as to the basis and general principles
upon which the system rests, no such classification, even were it possible,
would be universally accepted. Race, consanguinity, function, creed and
policy cover respectively a considerable portion of the ground, but no one
of them covers the whole or can be made the standard by which the
divisions as they now exist can be graduated on the social scale. It might
be thought that in view of the extreme value attached to conventional
purity, and the minute rules in regard to it by which the intercourse between
the different sections of the community is, by unanimous public opinion in
each locality, undeviatingly regulated, a touchstone might be found in it
by which social rank might be assayed. This, however, is not the case.
Irrespective of the difficulty of obtaining a formal decision on individual
cases, owing to prejudice and the general ignorance of the position of
classes below them which prevails amongst those who would ordinarily
be consulted, there is a marked difference in practice in regard to inter-
communion between the greater part of Continental India and the Peninsula,
and even between province and province. The criterion which would be
adopted would be whether or not certain higher classes would take from
the community in question water or certain kinds of food, and these lines

Social Organisation. B. Descriptive. 23

of demarcation are in most cases so far apart, including that is, so many
communities in each class, that they afford little or no graduation of the
masses respectively enclosed within them, and without further internal
subdivisions the groups are of little practical significance. Now, for the
purpose of this review, which is mainly to render the facts assimilable by
those who have not been brought into personal contact with the civili-
sation of India, the basis of that subdivision will be found in function,
overlying in some cases a distant but traceable background of race. It
will be found that, as a rule, graduation upon this basis is in general
harmony with the current conceptions regarding hereditary- puritj' which
prevail in India. The term function, it should be explained, is not limited
to the occupation actually followed in the present day, but extends to that
traditionally ascribed to the body in question, and is more frequently than
not implied in the title of the caste. This expansion of meaning is neces-
sitated by the mobility of occupation in modern times, on the one hand,
and, on the other, by the consideration that whilst function usually takes
rank in relation to purity according to the character of the service per-
formed or of the material handled, there are numerous cases where the
public estimate is formed upon the origin of the community by whom the
occupation is pursued, and thus takes its stand upim racial considerations
rather than upon the intrinsic nature of the pursuit. Elsewhere, again, race
alone is the determining factor; but here the community, as a rule, stands,
as explained above outside the Brahmanic system. The influence of the
latter, however, extends far beyond the limits of the Brahmanical religion.
The definition of caste quoted above is therefore applicable without serious
modification of its essentials to communities of not only Jains and Sikhs,
but, except in the North-west, even of the Muslim persuasion, as they rise
in wealth and in the power which wealth, even under Brahmanism, is able
to exercise. These instances have been included, accordingly, in the review
which follows, important differences of religion being duly noted against
them. As regards the review itself, it is not intended to serve as a Glossary-,
or to give an account of all the castes and tribes which find place in the
Imperial returns, but merely to bring to notice the principal bodies under
each of the heads into which Indian society has, for the purpose of exposition,
been here marshalled on the lines laid down above.

§ 23. At the head of the list are placed certain groups of an exceptional
character, whose position differs somewhat from that of the rest. The Brahman
naturally stands first, as the keystone of the whole social scheme. The
Rajput, again, is an order of nobility rather than a caste in the ordinary
acceptation of the term. With, but after, them may be taken the trading
and writing classes, both of which in Upper India, though not in the South,
claim distant connection with the Rajput, and who, with the Brahman, con-
stitute what are known as the Educated classes of India. Here, too, may
be placed the religious devotee, or mendicant orders, who, by virtue of
their profession have abjured caste, though in more than one instance only
to re-form themselves into something very like a caste of their own.

In dealing with the masses of the population, the first fact of which
cognisance should be taken in regard to the general arrangement of the
castes is the remarkable preponderance of the agricultural clement. Culti-
vation is the premier employment of the country, and to occupy a holding
is the main object of the bulk of the rural population. In the little oligarchy,
therefore, known as the village community-, the landed classes stand at the

24 5- Ethnography.

top, and where, as in all but the east of India and the tracts still under
the forest tribes, that community exists in an organised form, the classes
included therein are all subservient to the needs of the peasantr)'. Each
of these economic units contains a recognised body of artisans, minor pro-
fessionals and menials, to whom is assigned respectively a small share of
the village land or of its annual produce. Mixed in with these, are found
the various large bodies of fishers, cattle-breeders and others, some of
whom hover between the fields and their eponymous means of subsistence.
To the village, then, as it is understood in India, is dedicated the second
of the main divisions of the list, followed by a small group of minor, or subsi-
diary professional castes between village and town. In the third are placed
the castes exercising functions specially <jr exclusively the product of city
life. In placing them after the rural bodies it is not implied that they rank
below the latter from whom in most cases they originally sprang, for they
stand, as a rule, a little higher; but they are, as it were, bye-products of
the hive, outside the normal output, and on lines parallel to the main
organisation. Then, detached from either town or village, except in a few
cases where a permanent pied a terre is kept for shelter during the
rainy season, are various tribes of travellers and nomads, some of whom
are real castes, others a nondescript collection of waifs often consisting of
"broken men" or people discarded by other communities. The greater
number of the latter are numerically small; but there are a few which include
large and respectable communities. Finally, some reference must be made
to the bodies not coming within the caste system, such as the more or
less primitive tribes of the Hill-tracts, and also the Muslim races foreign
to India in their titles, though to a great extent native in blood.

As regards the arrangement of the items coming under each head,
functional or other, it seems best to deal with the return territorially, or
by linguistic divisions, as the case may be, in order that prominence may
be given to the marked differences in the caste system which are found
to prevail.

CASTES AND CASTE-GROUPS.

A. Special groups.

§ 24. Brahmans (14,893,300). Considering that the participation of
a Brahman is essential to the validity of all ceremonies of a social cha-
racter amongst the great majority of the community which takes its religious
title from this order, it is not surprising that the latter should occupy the
first place in the returns both as to numbers and dispersion. In every
part of India, except the eastern and western frontiers and the hills of
the Central belt, the Brahman is found in very considerable numbers, and
tradition, which in this case, at all events, is corroborated by the evidence
of physiognomy, nomenclature and custom, is almost unanimous in pointing
to the upper Gangetic region as the place of origin. From this nucleus
Brahmans found their way in very early days across Rajputana and Malva
to the west coast of Gujarat. In the south of the Peninsula, the earliest
appearance of this class was probably not much earlier than the Christian
era, and for the next eight or nine centuries the supply seems to have
been plentiful and constant. The Brahmans of lower Bengal trace their

Castes and Caste-Groups. A. Special Groups. 25

origin back to the loth century, when a considerable colony was imported
by the reigning sovereign from upper India and acclimatised in the north
and west of the present Province. Orissa received, or produced, its stock
a little later, but there seems some reason to think that there was an
earlier strain which had become extinct, or had degenerated below the
standard exacted by the dynasty which had established itself on the coast.
The frequent invasions of upper India from the north-west during the
ten first centuries of the Christian era are credited with the dispersal of
large bodies of Brahmans from Rajputana and the JMadhyades'a, some of
whom took refuge in the seclusion of the Nepal valley, others in the
west Dekkan ; others, again, fled by sea through Sindh or Kathiavad to
various settlements along the west coast. Amongst the latter were at least
three Brahman communities who have preserved a credible tradition of
their northern origin. The Brahman was never organised into a tribe upon
a territorial basis, but was, from the beginning, parasitic upon other classes
of the community. In Vedic times he was part and parcel of the fortunes
of the Chief, his patron. In later times, as the tribes settled, multiplied
and expanded, he attached himself to the landed classes, his principal
clients, for "unde vivent oratores si defecerint aratoresr" Still later, again,
he was liable, according to numerous traditions current amongst the
Brahmans of to day, to be imported in large bodies to a distant Court
on the invitation, not always declinable, of the pious ruler. When, moreover,
there is taken into consideration the incorporation into the Brahmanic
order of local communities and of priests and exorcists of the wild tribes
accepting Brahmanism, the capricious exercise of the powers of Brahma-
nification arrogated to themselves by sundry of the Chieftains, and the
results of left-handed unions with the daughters of the land, the extent
to which the Brahman is scattered far and wide is no matter for surprise.
The land, however, where they first became a consolidated body and
established the hierarchy they have since dominated, is still that in which
their numbers are both absolutely and relatively the greatest. Between
the Jamna and the Ghogra, roughly speaking, there are about 4S00000
Brahmans. Of the vast population of Bengal, 2900000 are of that order;
these two Provinces, therefore, account for more than half the total number.
Brahmans abound, too, relatively to the population, in Rajputana, and
Madras, Bombay and the Panjab each contain between a million and a
quarter. The distribution over these large areas is not, of course, even.
Orissa and Bihar stand out above the rest of Bengal, except for a few
places in the centre of the Province. Further up the Ganges, Oudh sur-
passes the sister Province of Agra in the relative number of its Brahmans,
and it is worth noting that Gonda, the traditional seat of the Gaur section
of Brahmans still maintains its preeminence. The prevalence of Brahmans
along the eastern bank of the Jamna extends also for some distance to
the west in both Rajputana and the Panjab. In the former tract there is
a large settlement in the so-called desert States of the north and west,
but in Sindh and towards the domain of the Baluch and Pathan, scarcely
any are to be found. In the Panjab, the greatest relative prevalence of
the sacerdotal element is found in the outer-Himalaya, where Brahmanism
reigns in unwonted vigour. In the west of India, the Brahman is well re-
presented on the wealthy plains of Gujarat, and holds a strong position
throughout the Dekkan. In the Dravidian tracts, his numbers are fairly
evenly distributed over the main linguistic divisions.

26 5 Ethnograpfts'.

§ 25. In spite of the unique and universally recognised position the
Brahmans hold in the estimation of the multitude, they have never formed
themselves into a single and homogenous body. Their very dispersal over
the length and breadth of the continent, in communities different in origin,
speaking different languages and eating different food, makes such co-
hesion impracticable. It has, indeed, had the effect of making them perhaps
the most heterogeneous collection of minute and independent subdivisions
that ever bore a common designation. Possibly, too, the absence of terri-
torial settlement to which reference was made above, lends greater weight
and permanence to a subdivision based on considerations other than those
connected with landed property, and has promoted, accordingly, the stricter
observance of caste separatism. However this may be, the main lines of
distribution are geographical, beginning with the ancient partiti<m of the
Brahmanic order into the five Gauda, or Northern sections, and the five
Dravida, of the South. To the former belong the Gaur, frt)m Gonda in
Oudh, the Kanaujia, of the Central Doab, the Sarasvata of the upper
Jamna, the INIaithila, of Tirhut, and the Utkala of Orissa. South of the
Vindhya come the Maharastra, of the Dekkan, the Karnata, of Mysore
and the neighbourhood, the Andhra of Telingana and the Dravida of the
Tamil country. Added to these are the Gurjara of the west, who, curiously
enough, though grouped amongst the southerners, are all northern in their
origin. Except in the case of the three first mentioned, these divisions
are of little practical significance in the everyday life of the present time,
since they are severally partitioned into numerous main subdivisions, each
of which is in turn, again, minutely split up into a still greater number
of separate endogamous communities. The majority of the larger castes
thus constituted have a territorial origin, generally well to the north of
where they are now settled, except, of course, amongst those still occu-
pying the traditional centres of Brahmanism, such as the Gaur, Kanaujia
and Sarasvata. Subordinate to these are the local offshoots, which are verj-
generally attributable to schism on points of ceremonial or food, and, in
the Dravida country, to sectarian or doctrinal disagreement. From time
to time, too, the scheme has to be expanded to admit some new recruits
from outside the fold, who are usually placed low down on the scale,
though not irrevocably doomed to remain there, if circumstances turn
out favourable to- their advancement. Throughout the local community,
the rank of each subdivision relatively to the rest is fixed by a convention
effectively backed by the public verdict; but this graduation is not neces-
sarily recognised at a distance or where a different language is spoken.
In every linguistic grouji, moreover, there are certain classes which, though
called Brahmans by the public, and enlisted to perform st>me of the ce-
remonial functions of the Brahman, are either not recognised by other
Brahmans, or are relegated by them to a degraded position, inferior, in
reality, to that to which many of the non-Brahman castes are admitted.
The acme of subdivision in combination with ceremonial exclusiveness,
is probably reached among the Kanaujia, of whom it is said in their native
Province, "Three Kanaujia, thirteen cooking-fires". The Gurjara Brahmans,
again, are popularly credited with S4 divisions, but this being a popular
expression of multitude in general, the number actually found, viz. 79, may
be taken as fairly correct, especially as all the larger items in that lengthy
list have their respective sub-castes. The Brahmans of the Dekkan are
perhaps as little split up into sections as any, but on the coast-strip of

Castes and Caste-Groups. A. Special Groups. 27

the Konkan the subdivision is more minute, owing, probably, to the foreign
strain introduced from time to time. The Brahmans of Bengal and Madras,
where the system is of later introduction, hide a complicated interior under
a comparatively small number of main divisions, especially in the latter,
where caste has been affected by the doctrinal schisms of which the clouth
since the days of S'aiikaracarya and Ramanuja, has been prolific.

§ 26. It was stated above that the subdivision of Brahman communi-
ties is often traceable to differences in regard to food and ceremonial.
These, in turn, depend to a great extent upon function and the means
of livelihood accessible. Strictly speaking, the Brahman, as pointed out
earlier in this work, is by origin a functional order, but with the ex-
pansion of the Arya population in post-Vedic times and the growth of the
Brahmanic ct)mmunity beyond the need of the layman for its specific
ministrations, great latitude had to be allowed, no doubt, from a very early
period. In the present day, within the fairly wide limits which he himself
has set, the Brahman is represented in a large proportiim of what may
be called the upper and middle class occupations of India. But whichever
of these he may take up, his inherent qualities are unabated, and he is
still entitled to the homage of the rest of the community, and remains
the accredited intermediary between man and the supernatural. In the
latter capacity his bare living is assured to him without need to work for
it, because in all formal rites such as those connected with birth, marriage,
death, expiation or thanksgiving, the provision of a meal for a certain
number of Brahmans is an essential and costly feature. In the more pros-
perous parts of the country', accordingly, there is usually a plentiful supply
of Brahmans of whom it has been said that "they exist only to be fed".
On every side are to be found subdivisions which, in the eyes of their
compeers, have fallen from grace by participating in the feasts of wealthy
but impure clients. In another direction there are instances on record
where the number of local Brahmans available for a ceremony of this
sort not being equivalent to the aspirations of the Chieftain interested in
it, the quorum has been made up by him by a special creation out of
such lower material as was at hand. Service at a temple, it should be
noted, is not undertaken by the better class of Brahman, as it is held to
be degrading, and left, accordingly, to those low in station. In several
cases the claim to be accounted a Brahman rests entirely upon the per-
formance of those duties. The inference drawn from this estimation of
temple service is that the divinities in question are those of the non-Arya,
incorporated from time to time into the Brahmanic pantheon, as the com-
munity which reverenced them was brought to adopt the social system
of the higher race. It is probable that the distinction drawn between the
acceptance of offerings by a Brahman in requital for specific services and
those made to him on general grounds has its root in the same tradition;
for whilst to the donor offerings of any kind to a Brahman are held to
be productive of spiritual merit, only one of the lower class of the order
will accept gifts for exorcising evil spirits, averting the baleful influences
of an eclipse or certain combinations of stars, reciting the appropriate
texts for pilgrims at a bathing place, or helping at a funeral, and the
like offices.

The secular pursuits affected by the Brahman vary considerably ac-
cording to whether the caste is settled in the locality in large numbers,
whether the tract is prospering, or whether the Brahman first came into

28 5- Ethnography.

it as a pioneer and colonist or as a propagandist or an exile from another
centre. Political employment has been congenial to the Brahman from the
time when the Purohita, or family sacrificer, was treated by the Rajan
as his confidential adviser in the Sukta period, and the caste has con-
tinued to throw up from time to time men who have been distinguished
for their administration of Native States. The great chance of the Brahman
came, of course, under the Pcsva rule, when the whole of the military
organisation built up by the Marathas fell to the disposal of the Citpavan
Brahman of the Konkan; and for seventy years or more, the Dekkan was
dominated from Poona, and the whole of the administration was conducted
by the local and the coast Brahman. Even in the present day, the Maratha
Brahman has almost a monopoly of clerical employment throughout the
Dekkan, Konkan and Karnatic, and with the traditions of former supre-
macy to encourage him, he stands quite in the van of his order in in-
telligence and general ability. In some other parts of the country the
Brahman is the only class besides the trader who can read and write to
any practical purpose, and he thus becomes, of course, the scribe, if not
the official accountant, of the village community. Even in the tracts where
a serious rival is found in a professional writing class, the Brahman usually
has a share in the State appointments to which the "literary proletariat"
of India look mainly for their subsistence. Of the learned professions. Law
and Instruction arc the more attractive to this caste. A few take up the
lower branches of Engineering, and still fewer the practice of Medicine,
a following which is to a great extent barred to them by reason of caste-
scruples in regard to the surgical training involved. In commerce they
have not made their way beyond the universal venture in lending money
to their neighbours, to which every Indian capitalist, according to his
resources, is inclined. The Brahman shares, also, the general aspiration
to own land, either as an investment or as a possession honorific in the
eyes of the lay world. Wherever they have settled in large masses, as in
the Gangetic Doab and Oudh, or in compact local colonies, which pro-
bably preceded their advance as a sacerdotal body, they have taken to
cultivation on the same lines as the ordinary peasantry, except that they
but very rarely put their hand to the plough, though they go as far as
standing upon the crossbar of the harrow to lend their weight to that
operation. Owing to this caste-imposed restriction, probably, it may be
noted that wherever the Brahman has settled otherwise than as a part of
a large general community, he is the centre of a well-defined system of
predial servitude, his land being cultivated for him by hereditary serfs of
undoubtedly Dasyu descent. This is the case with the Masthan of Orissa
and Gujarat, and with the Haiga or Havika of Kanara, and the Nambutiri
of the Malabar coast, all of whom have settled in fertile country. Where
the pressure of circumstances is very severe, as in the desert States of
Rajputfina, ihe Brahman cultivator not only does the whole of his own
work, including ploughing, but even sells his labour to other more fortunate
occupants. A military career may appear to be somewhat alien to the tra-
ditions and inclinations of a sacerdotal class, nevertheless in the vicinity
of the Ganges it has proved by no means unattractive to the Brahman
peasantry. The Bhuinhar, or Babhan, of the south eastern parts of the
upper valley, are credited by some with Brahman ancestry, which endo-
wed them with enough of the Ksatriya qualities to enable them to push
forward in advance of the main body of their race, and to hold against the

Castes and Caste-Groups. A. Special Groups. 29

Kol, or other previous possessor, the land they still occupy. The nickname
of Pandy, again, bestowed upon the rebel troops collectively by the
British soldier, is no other than Pandc or Pahre, the title of the sub-
division of the Kanaujia Brahmans from which a high proportion of the
recruits of this caste were then enlisted. Since 1S57 it has been found
that minute caste-scruples as to diet and contact arc incompatible with
the exigencies of modern field service, whilst the personal sanctity of the
Brahman private is apt to turn out inimical to the due observance of re-
gimental discipline. Amongst the Muhial Brahmans of the Panjab, therefore,
the pride of caste has given way before the taste for the profession of
arms, and the would-be recruit of this community drops his Brahmanhood
when enlisting, and is enrolled under some other designation. In former
days, when, as under the Pcsvas, Brahmans were themselves at the head
of the forces, and not in subordination to the foreigner, and when war
was carried out on very different lines from those of to-day, commanders
of this caste acquitted themselves worthily, and showed both resource and
courage in the field.

!< 27. Rajputs (10,040,800). In this case, the community is unmis-
takcably military in its origin, with the old baronial attributes of landed-
estate and leadership of an armed force. People are returned at the Census
under this designation in considerable numbers from all parts of India
except the South, but nine-tenths of them hail from north of the Vindhya
and west of the Kosi. The Provinces of Agra and Oudh alone account
for 3,950,000. In the Panjab there are 1,820,000, and in Bihar, about
1,200,000. The cradle of the Rajput is the tract named after him, not,
however, as it is limited in the present day, but extending from the Jamna
to the Narbada and Satlaj , including, therefore, the whole of Malva,
Bundelkhand, and parts of Agra and the Panjab. From the northern parts
of this tract there seems to have been an early movement of conquest
up the western rivers of the Panjab, as far as the Himalaya and Kashmir,
whereby was laid the foundation of the predominance of the tribes still
in possession. With this exception, the presence of Rajputs in other parts
of India seems due to their expulsion from their ancient seats. The le-
gendary occupation of Kathiavad from Mathura is ascribed to an attack
delivered from the south and east. Successive inroads of Scythians and
Hunas caused a movement to the south-west, into Gujarat; but the prin-
cipal and most definite migration followed upon the Muslim conquests
of the nth and 12th centuries, which drove large bodies of Rajputs to-
wards the Himalaya and eastwards across the Ganges into the Doab and
Oudh. From thence, as well as from Bundelkhand, they spread into the
adjacent parts of Bihar, especially those north of the Ganges. A certain
number, too, are found in the north of the Central Provinces, where the
boundaries between British territory and Central India are very ciimpli-
cated. Beyond the above limits the original stock is Udt found, and even
within them, it has in some cases been materially watered with local blood,
when the distance from the race-centre makes the operation fairly safe,
and the community is sufficiently well established to maintain its marriage
connection at its conventional level. The presence of so many Rajputs
in other parts of India is accounted for by the fact that the title, originally,
in all probability, derived from function, denotes, as has been stated, an
order of hereditary nobility, access to which is still obtainable, and whose
circle, accordingly, is being constantly enlarged upon much the same con-

30 5- Ethnography.

sideratiuns as of yore. The essentials of the position are the chieftainship
of a tribe or clan and the command of an armed force, with the possession
of a substantial landed estate and a scrupulous regard for the strict letter
of Brahmanical regulatiijns as to marriage, domestic customs and inter-
course with other classes. It was on this basis that in the Panjab the Jat
was differentiated from the Rajput, and certain castes in Oudh and its
neighbourhood rose above their fellows. In other cases, the above re-
quisites being established, the elaboration of the claim u> affiliation to one
of the recognised Rajput clans is left to the ingenuity of a competent
Brahman with the aid of an experienced bard or genealogist. For example,
on the adoption of Brahmanism by a large portion of the Mongoloid po-
pulation of Manipur, the chief and his military retainers passed into the
rank of Ksatriya, and to the number of about iSoooo, appear under that
title in the last Census returns. The leading families of various K61 tribes
of Chutia Nagpur, again, are constantly, in Col. Dalton's phrase, "being
refined into Rajputs" and sometimes do not wait for 'times' effacing
fingers" to conceal the change, and too often ignore the essentially
Rajput system of clan-exogamy in favour of their pristine tribal arrange-
ments. There is, in fact, no section of the Brahmanic hierarchy into which
recruitment from the outside has been more extensive or to which the
claims to membership have been so numerous. The latter is especially
tHfe case in the tracts where the caste system has been imposed as an
exotic in comjiaratively modern times. In Lower Bengal, for instance, such
claims are remarkably frequent, and this is attributed to the adoption of
the ready-made caste-system by a number of different racial stocks without
its graduation being authoritatively regulated by a powerful Chief under
the guidance of a council of influential Brahmans. In Madras, again, caste
was engrafted upon an already well-established civilisation to which it had
to accommodate itself according to circumstances. In the former, therefore,
the Rajput, except as before stated in Bihar, is redolent of the local soil,
and takes rank therefore below certain other castes which have come to
the front under the peaceful conditions of a Province where arms have
long succumbed to the tongue and pen. These classes, therefore, do not
lay claim to the title of Rajput, but to that of Ksatriya, implying a po-
sition less definite and less likely to be disputed by existing ct)mmunities.
Similarly in the South, whither the Rajput never penetrated, unless it might
be in the form of representatives of more or less evanescent dynasties,
the rank of Ksatriya is claimed almost exclusively by members of the
labouring and toddy-drawing castes, who justify their pretensions by the
undisputed fact that their ancestry furnished the rank and file of the
archers and other infantry of the local potentate. Instances will be found
in latter parts of this Chapter in which the status of Ksatriya is claimed
by many castes of far higher position in the present day than those just
quoted. Various legends are current proving that whilst the Puranic as-
sertion of the total extirpation of the Ksatriya is true, the ancestry of
the claimants in question had somehow or other escaped the general
destruction, and are the lineal inheritors of the hypothetical Vedic rank,
although the majority of them obtrusively avoid any occupation savouring
of war. This much appears to be true, that there was a long breach
between the heyday of the post-Vedic ruling classes and the genesis of
the Rajput. The former were apparently staunch supporters of Buddhism,
in its inception a movement in their favour, whilst the latter arose with

Castes and Caste-Groups. A. Special Groups.

the forces which deposed that religion in India, and established their po-
sition upon the ruins of the States which had professed it. The ground
for the evolution of a new military nobility seems to have been prepared
by the establishment in Upper India of successive sovereignties of S'aka
race. These professed Huddhism, and were thus antagonistic to the
orthodo.x Brahmanism. But after they had carried their arms far into the
country, and the Panjab and its neighbourhood became their principal
seat of government, they seem to have become affected by the prevailing
social atmosphere, with which, as has been stated, the tenets of Buddhism
were by no means out of harmony. One of their monarchs, indeed, is
claimed as their founder by more than one of the chief clans of the
present-day Rajputs. In the continual disturbances which occurred between
the first century before Christ and the downfall of the principal Scythian
dynasties in the 7th century A D, the Brahmanic powers were wont to
invoke the aid of any arm, Indian or foreign, which might promote the
defeat of their rivals. The incorporation of such leaders into their ranks
could be effected without much difficulty, firstly, through the prestige of
a victory in the good cause, and, again, through the fiction, dating from
a far earlier period in Indian history, that the foreign tribes which pressed
upon the frontiers of Brahmanism were themselves Brahmanical back-
sliders of the warrior order, who had lost their position by reason of their
neglect of the orthodox rites. Upon the hypothesis that the suppression
of Buddhism was an act of faith entitling the protagonists to be received
back into the fold, it became possible to combine gratitude with policy,
and, by the substitution of a new designation, Rajput, for the old one of
Ksatriya, to effectively demarcate from the former state of things, the
new order established under the uncontested supremacy of sacerdotalism.
None of the Rajputs prove their pedigree further back than the 5th century
of the Christian era, and four of the leading tribes of the present day,
known as the Agnikula, or Fire-clans, derive their origin from a specific
act of creation under Brahmanic auspices, whereby the sun and fire-
worshipping Hijna or Gurjara was converted into the blue blood of Raj-
putana, and became the forefathers of the Sisodia, Cauhan, Parmar, Parihar,
and Solariki or Calukya, and perhaps of the Kachvaha lines. Other cases
of similar elevation are to be found, and, considering the dominant position
held by Scythian communities in the north and west of India for many
centuries, together with the affinity between their worship and that of a
popular branch of that of the Brahmans, and the common northern origin
of the two races, it is not improbable that the upper classes, at all events,
of the new comers should have identified themselves with the correspon-
ding classes of those amongst whom their lot had been permanently cast.
There are, moreover, special features of the structure and customs of
Rajput and Jat and other northern communities in India which distinguish
them from the Brahmanic masses of the interior, and may be attributed
to difference of race, perpetuated by many generations of resistance to
attacks from the outside. The least that can be said is that a race-con-
nection of the above description could not possibly have existed so long
and then faded out without leaving substantial traces of its passage upon
the people subject to it. It may be added that Rajput dynasties did not
rise to power until sometime after the Hiina supremacy had been broken
in the 6th century, and that the genealogies of the tribes now ruling
States start from about the 7th century. The contests with the Muslim

32 5- Ethnography.

invader of a few centuries later had the effect of consolidating the Rajput
devotion to the scrupulous observance of Brahmanic injunctions as to
marriage and intercourse with other castes which specially distinguished
them from their foreign oppressors ; and to the present day, they stand
out from the rest of the community in the high value they attach to these
matters. Like the Brahmans, they are greatly subdivided, but with this
important difference, that whereas the Brahmans may only marry within
the subdivision, the Raj])ut may only marry without it, though within the
Rajput pale. The larger subdivision is, in fact, taking the place of the
smaller as the circle of prohibited affinity. Conjecturally, this difference
in practice may be due to the fact that the Rajput clan is definitely tra-
ceable in its origin to a historic leader or family, involving, therefore, a
tradition of blood-kinship the more vivid from its being associated with
territorial ownershi]). The tribe or order, again, being spread continuously
and in considerable numbers over a large area, with uniform conceptions
as to rank and function, the marriage field is a wide one, and the gra-
duation of each unit in its social position has been arranged on conside-
rations which override the normal limitations of caste. The regulations as
to intermarriage, therefore, though exceedingly strict, have a wider scope
than among most of the other Brahmanical bodies and are in some cases
arbitrarily imposed upon itself by the clan on considerations of rank alone.
So strict indeed, arc they in regard to what has been called hypergamy,
that amongst the ujiper grades of Rajput society, the girl is held to be
a burden upon the resources of the family to an extent that leads to
reprehensible means of preventing her from reaching a nubile age. The
scarcity of brides thus produced, combined with the expenses of the
marriage, tend to the formation of left-handed unions with lower castes,
the offspring whereof ranks with the mother, or, where numerous and
recognised, constitutes a new caste by itself. The latter is the case in the
west of India, where the bastards become court dependants. In Orissa,
they all rank as Rajputs. In Nejial there is the curious instance of the
children of a Hill woman by a Brahman becoming Rajput, and forming
the kernel of the large military population of the State. In the Kangra
Himalaya, where the continuity of tradition and lineage has been less
interrupted than anywhere else, the Chief is a law not only unto himself
but unto his subjects in regard to social position and caste, so that the
rank of Rajput depends very much upon the royal favour. Considering
the part played by Islam in the dispersal of the Rajput ruling families,
it is worth noting that in the Panjab, not only have three fourths of this
caste embraced that religion, in both the west and east of the Province,
but that conversion has had no effect upon the social position of the
Rajput. In the east, where Brahmanic influence is supreme, change of
religion is said to have no result upon caste regulations. In the west,
where the Pathan atmosphere predominates, the scheme of social restric-
tions and prescriptions is Brahmanic, but, as in the east, the sanction by
which it is maintained is that of the tribe, not of the caste, and inter-
marriage and so on is governed by the position of the body in the present
day, rather than by considerations of origin, such as are involved in caste.
From what has been said above it may be inferred that the func-
tional scope of the Rajput is but narrow. Traditionally, he rules, fights,
owns land and indulges in field-sjiorts. In practice, he carries out this
scheme of life as far as circumstances allow, but the rank and file of his

Castes and Caste-Groups. A. Special Groups. 33

order are cultivators, and not among the most efficient of their class. The
Rajput has the same objection as the Brahman to handling the plough,
and the strict seclusion in which the women of the caste are kept deprives
him of an aid in the minor agricultural operati(ms which in the lower
castes is often most valuable. In the Gangetic regions the Rajjiut still
enlists in considerable numbers in the "Hindustani" regiments of the British
army. He often, too, dons the official belt as a constable or messenger,
in upper India and Gujarat. On the whole, however, the general disincli-
nation of the caste to avail itself of the facilities for instruction now within
its reach is placing it at a disadvantage as compared with the middle
classes, in the modern conditions of Indian life. Only the subordinate grades
of official and professional employment are open to them, and in the army,
also, promotion beyond a certain rank depends now-a days upon educa-
tion, and the Rajput is losing by the competition of Sikh, Palhan and
Gurkha in the profession of his choice, and is far outdistanced in civil
avocations by those whom his caste prevents him from acknowledging
to be even rivals.

ij 28. Trading Castes (io,6So,8oo): This is the first of the distinctively
functional groups to lie brought under review. It is not, however, merely
a collection of communities each with its separate designation, like those
which have preceded it, but contains some general titles denoting the
occupation of trading, but which do not include all the castes following
that calling in the locality where it prevails. The leading example of these
is the Banya or Vania, of upper and western India, under which name
are included nearly all the trading classes, but not important castes like
the Khatri and Arora in the Panjab, or the Bhatia and Lohana of Sindh.
There are grounds for thinking that the exclusion is due to differences
of race. The Khatri and Arora, like the Banya, derive their origin from
Rajputana, in the larger sense in which that term was used in the preceding
paragraph, but the latter affiliate themselves directly to certain clans of
Rajputs, whilst the former refer themselves back to the Ksatriya, and
give the western region of Multan and upper Sindh as the cradle of their
caste. It has been conjectured from the customs and internal structure of
the Khatri and Arora, which differ in some respects from those of the
ordinary Brahmanic castes, that these communities are descended from
one of the S'aka colonies which long held the tracts above mentit>ned.
The Banya, with the exception of the Agarval, who come from Agar near
Ujjain, give the now ruined city of Bhinmal, or S'rimal, in Marvad as
their original home, and claim descent from the Solanki clan of the
Agnikula or Huna Rajputs, so that, like the Khatri, they are of foreign
race. Whether owing to this origin or to the refining influence of gene-
rations of sedentary pursuits in prosperous circumstances, the personal
appearance of the Banya is decidedly above the average. The western
subdivisions, such as the S'rimall, Porval and Osval, which are all closely
connected with each other, are largely, and in many tracts, mostly, of the
Jain religion, a creed which seems to have commended itself to the mer-
cantile community at a comparatively early period ; and they allege the
acceptance of the peaceful tenets of this faith to have been one of the
main reasons for their separation from the bellicose Rajput. In the present
day, except in Delhi, where a special casus belli arose some years ago,
the Mahesrl, or Brahmanic section of the caste intermarries with the S'ravak,
or Jain, and the latter, in turn, employ for their caste and domestic mi-

InJo-Aryan Research. 11. 5. 3

34 5- Ethnography.

nistratiuns, the Bhojak, or Scvak, a subdivision of Brahmans not in high
repute among the priestly orders, representing, as they are said to do,
the Maga sun-priests introduced from Iran l)y the Huna and other invaders.
In addition to the main divisions of the Banya, almost every body is sub-
divided into "full-scores" (visa) and "half-scores" (dasa), denoting the
relative admixture of lower blood. In many castes the partition has to be
carried still further, and the "quarter-score" (pafica) represents the
minimum of pure descent. None of the subdivisions intermarry, though
in the west there is occasional connubium found between the "visa",
or highest sections of the respective castes. The Banya engage in most
mercantile pursuits, from high finance and extensive fureign trade down
to the retail of the most common articles of everyday use, so long as
these are not conventionally polluting. They are not as wedded to their
native place as most of the Indian communities, and settle, sometimes
permanently, in villages where they are strangers both in caste and
language. Others, jirincipally from the desert States, habitually leave home
for the more favoured parts of t,hc country, and return only after their
fortune is made there. The upper classes of the Banya are well educated
and often keen sectarians in regard to religion. In some tracts they are
entering the law and the State offices, though not in large numbers. The
Khatrl of the Panjab, on the other hand, in addition to the trade of all
but the south-west of his province, has almost the monopoly of official
and professional employment, and has passed even beyond the Panjab
into parts of the neighbouring jjrovince in similar callings. This caste has
what the Banya lacks, the tradition of administrative and political success,
in which it resembles the Maratha Brahman mentioned al)ove. Todar Mai,
the celebrated financier under Akbar, was a Khatri, and has had more
than one successor, though not of the same calibre. Then, too, though
the bulk of the Khatri are not of the Sikh faith, they have always been
connected with it, and both Nanak and Govind belonged to their ranks.
In the present day, such priests as are required by the Sikhs are usually
Khatri. In trade, though sharp and industrious, the Khatri does not take
so high a position as the Banya, but confines his operations generally to
small local transactions, and does not, as a rule, set up branch establish-
ments outside his native province. There are, however, a few colonies
in Bengal, but they are detached, and their position is considerably below
that occupied by the caste in its northern home. In some other parts of
India there are Khatri returned who trace their origin back to the Panjab
or north Rajputana, and were probably driven southwards by one of the
Scythic cataclysms, and like others similarly circumstanced, found them-
selves obliged to take to new means of livelihood, generally silk-weaving.
Closely allied to the Khatri, but occupying a decidedly inferior social
position, are the Arora of the south-western Panjab, who, starting from
nearly the same region as the others, do not appear to have pushed their
way into the fertile tracts of the north, but to have remained on the less
remunerative plains along the Indus. In the same direction are the Bhatid
and Lohana of Sindh. The former have preserved in their title the me-
mory of their origin in the Bhatti districts of north Rajputana, and claim
descent from the predominant Rajput stock of that locality, just as the
Banya of Bhmmal does in the west. There is this further similarity, that the
Yadava race of the Bhatti looks back to a S'aka founder, in the grandson
of Kaniska. There are still a good many Bhatia in the Panjab, where their

Castes and Caste-Groups. A. Special Groups. 35

Rajput blood seems unquestioned, but, unlike the Khatri, their position
seems to improve the further they get from their native country, and it
is along the coast that they are most flourishing, and in upper Sindh most
depressed. They have so arranged their caste-rules that they are able to
cross the ocean without subsequent trouble, and are among the most
travelled and enterprising merchants of Kach, Bombay, Zanzibar and
even China. The Lohana, again, are of Marvad origin, but moved into
Sindh very early in their history, and have there remained. From a centre
at Shikarpur, they travel far into Central Asia and even to the banks of
the Volga. One of their subdivisions (the Amil) has followed the example
of the Khatri, and taken to clerical professions. Like its prototype, also,
it has succeeded in monopolising the pick of official employment in its
native province.

The figures given against the several subdivisions of the general
heading of Banya are much below the reality, owing to the omission in
many, if not most, cases to enter the subcaste, and to the substitution of
some such indefinite designation as Marvadi, S'ravak, Vais, and so on.
This is markedly the case in Bengal, where, with the exception of the
subdivisions dealing with specific products, which find place in a later
paragraph, nearly the whole trading community appears as a single item.
The Subarn abanik, it is true, has distinguished itself from its neighbours,
probably because it claims a rank above that accorded it by public opinion.
It is an immigrant body from upper India, and as a considerable number
of its members are still engaged as assayers and money changers and it
employs Brahmans recruited from its own ranks, it seems possible that
it is an offshoot of the Sonar caste which elsewhere in India makes similar
claims and is not unfrequently returned as a Daivajna or VLsvakarman
Brahman, an assumption not yet accepted beyond its own members. Other
artisan castes in the South make the same claim, but as the Subarnabanik
is prosperous and fairly well educated, it will not improbably end, if not
where it desires, at all events considerably above its present rank. In the
Dravidian country, the trading castes differ from those above described
in being almost entirely indigenous to the locality they serve. The move-
ments which are reported to have taken place have been to comparatively
short distances, such as those from the uplands of the Telugu country to
the rich and thickly-peopled tracts of the south-east. There is this further
difference between these castes and the traders of the north, that in most
cases the former are intimately connected with, and probably sprang from,
one or other of the great agricultural communities amongst whom they
live, and from whom they are still distinguished by little else than function.
One result of this relationship, and not an unhealthy one, has been observed
viz, that where the business of lending money is carried on by |)eople
of the same class as the borrower, the dealings arc on a less formal and
more elastic footing than where, as in other parts of the country, the
usurer has simply come to the village from a strange country to make
his fortune out of the necessities of the natives. Considering that what
with weddings and other ceremonies, every peasant is at some time or
other a borrower, the above feature is not unimportant from a political
as well as from an economical point of view. There is the usual tendency
among those who prosper to adopt the ceremonial and customs of the
local Brahmans or to grow more scrupulous in their observance, and,
amongst the Telugu traders, to assert in due course a Vaisya origin, a

36 5- Ethnography.

pretension which their form of caste-subdivision and their more intimate
domestic practices flagrantly contradict. The Komati, for instance, wear
the sacred thread and arc divided into three territorial endogamous sub-
divisions, lullowing the modern Brahmanic, not the Vedic, ritual. Their
exogamous groups, however, of which there are a great number, are
not Brahmanic but totemistic, derived from trees, plants or articles of
food, the use of which is prohibited respectively to the group to which
it belongs. Their marriage rules are those peculiar to the South and
the ceremony is incomplete without the formal presentation of the
friendly and symbolic betel -nut and leaf to a member of the impure
leather-working caste, with whom the traders share "a commim tutelary
deity. It is a good example of the growing refinement of modern times,
that in order to mitigate the crudity of the above-mentioned act of social
intimacy without breaking away from a possibly prophylactic tradition, it
is now the habit for the bride's father to send a pair of shoes to be
mended a few days before the wedding, and on the day of the ceremony
to pay the cobbler with a betel-nut thrown in to the amount of the bill.
The largest trading community of the Telugu country is the Balija, which
is widely sjiread over the Tamil di.stricts also, and there called Vadugan,
or Northerners, or Kavarai, from the caste goddess. They have a great
number of subdivisions, which are not, however, endogamous, as a rule,
possibly owing to the practice of receiving into the caste refugees from
outside who are in disgrace with their own kinsfolk. One division of the
Balija, however, keeps itself apart, being descended from the Nayak Chiefs
of Madura. Though it wears no sacred thread, it claims to be Ksatriya.
As a whole, the Balija arc probably an outgrowth of the great agricultural
body of the Kajiu or Reddi. Like the Komati, they are in curiously close
relations with the impure leather-workers and village menials of the lo-
cality. It might be inferred from this fact that the latter belong to a race
preceding the present occupants of the soil, and like the Dasyu of the
north, dispossessed of their heritage, but acknowledged to be influential
with the gods of the village. The Banjiga is the Karnatic trader, and has
no connection with his namesake the Banya. Generally speaking, the
Banjiga, though much subdivided, is of the same stock as the Kanarese
peasantry, whose ])roclivities towards the Lingayat faith it largely shares.
In the Tamil country the trader is usually a Cetti a title which is nearly
as comprehensive as that of Banya. It covers several large and a vast
number of small subdivisions. In most cases the marriage rules resemble
in imi)ortant particulars those of the surrounding peasantry of the better
class. They worship the local goddesses and call in a carpenter by caste
to bestow his blessing upon the bride and bridegroom, thus generally
testifying to their local origin. Their main subdivision, the Nattukottai,
shares the reputation of the Bhatia for unwonted enterprise and success
in foreign trade and travel.

§ 29. The last group to be mentioned under this head is that ol
the Muslim traders. These belong to the west coast, with the exception
of the Labbai, who, though settled along the south-east, are nevertheless
connected with those of Malabar by origin. The rest consist mainly of
converts of long-standing from the Lohana and other traders of Sindh
and Kach. I'nfortunatcly, the full strength of these bodies is not ascer-
tainable from the Census returns owing to the appropriation of the same
title by different communities. The Khojah, for instance, of the coast, are

Castes and Caste-Groups. A. Special Groups. 37

a wealthy body of enterprising traders converted to the Shiah form of
Islam about the 13 th century. They moved southwards from Sindh into
Gujarat and Bombay, and there, starting from petty shopkceping, they
have attained a very high position in foreign trade, and arc noted for the
number of the branches they have set up abroad. The Khojah of the
Panjab are quite distinct from these, though they too are converts from
the Brahmanic mercantile classes and mostly profess the Shiah tenets.
They also, like the others, recognise as their religious head H. H. Agha
Khan, whose family migrated to Bombay from Persia about sixty years
ago. The IMcman, again, arc of Sindh origin, descended from a body
of Lohana who were converted in the 15 th century, and, like the rest,
moved into Kach and Kathiavad. In common with the western Khojah,
they have preserved a good deal of their Brahmanic custom and tradition.
In commerce they have risen to a good position, though not, perhaps, to
the rank of the Khojah. Their counterpart is found in the Momin or
Momna, a body of Gujarat peasants converted about the same time as
the Meman, and who are now chiefly weavers and cotton-goods dealers,
with a few still on the land. About half of those returned as Mc-man at
the Census probably belong to the latter community, and about two thirds
of the Khojah are of the Panjab section. The last of the Muslim trading
classes of the Bombay coast to be here mentioned is the Bohra, in its
various subdivisions. These, like the rest, are converts to the Shiah faith
from the commercial classes of the chief towns in Gujarat, about the
1 1 th century, and combine the strict observance of Muslim worship with
a due regard for the Brahmanic or pre-Brahmanic methods of dealing
with the personal or domestic supernatural. The ujiper classes engage
in foreign trade, but the rank and file are content with a successful career
in the retail shop, and are somewhat remarkable for their neglect of
English in an otherwise efficient and well-diffused scheme of instruction.
The Census shows under the same title the cultivating Vohora of Gujarat,
Sunni by sect, and retaining in most cases a fairly clear recollection of
the Brahmanic caste from which they were converted, and adjusting their
marriage arrangements in accordance therewith. About half the number
of Bohras given in the return belong to this class. On the Malabar coast
are the Majjpila and Jonakkan, and on the south Coromandel coast, the
Labbai. The last named are descended from an Arab colony, driven from
its native country in the 8th century; or, according to another account,
from Arab traders who married Tamil wives at a later date. Their con-
nection with Arabia is indicated, in either case, by the name of S'onagan
(Arabia) which they used to bear, and their present name of Labbai is
said to be no more than a local rendering of labbaik, the Arabic for
the familiar phrase "here I am". In practice they are orthodo.x Muslim,
though like the Muslim of the eastern Panjab, they marry by Brahmanic
rites with a text or two of the Kuran recited to complete the ceremony.
There is a small community living side by side with them, known as the
Marakkayar, who claim similar origin, but do not intermarry, and are
apparently of more recent arrival. Both speak Tamil with a few Arabic
words interspersed. Those who are not traders are engaged in betel cul-
tivation and pearl-diving. The Mappila have been referred to in other
parts of this survey as the chief Arabian colony on the western coast.
They are placed in this group because it was as traders that they first
visited Malabar, but in the present day this pursuit is practised only along

38 5- Ethnography.

the coast, and the bulk of the Mapi)ila inland arc landholders and culti-
vators. In both capacities they have shown themselves thrifty and energetic.
Their name is cither an honorific soubriquet, shared by some other classes
in the neighbourhood, or, as some think, the Tamil word for bridegroom,
applied to the Arabs who married native women. In language and in many
of the local customs of marriage and inheritance, they have identified
themselves with the native population. The Jonakkan are no other than
Mnppila returned under a title given along the coast, especially in Tra-
vancorc, to converts to Islam, and is possibly the Malayalam rendering
of Yavana, the old Rrahmanic designation for all foreigners hailing from
the west. The community is recruited from some of the castes along the
coast, especially the fishermen, of whom the Mukkuvan have in some
families the curious rule that one of their children should embrace Islam.
In remarkable contrast to the experience in the Panjab in regard to such
conversion, it is alleged that the Malayalam is improved by the change
in faith. Probably the original status of the convert was lower than in
the north.

ij 30. Writer castes (2,750,300): The profession of scribe or clerk
was in all probability unusually late in establishing itself in India owing
to the jealousy with which all instruction was monopolised by the Brahmans,
as well as to the extraordinary development of memory and oral tradi-
tion fostered by them. Setting aside the art of inscribing rock and copper,
writing as a profession appears in inscriptions of the Sth century A. D.,
and a few generations later, the caste of the Writer is referred to under
the same name as it bears in the present day. It may be gathered from the
data available that the calling was in anything but good odour amongst
the Brahmans and that the castes exercising it occupied but a low position.
Their chance came when the Muslim conquerors, having established them-
selves permanently in the country, felt the need of clerical ability to
help them through the labours of administration, and were unwilling, on
sectarian grounds, to have recourse to the Brahman. In the writing castes
the very material they wanted was at hand. The Khatri, as mentioned in
a former paragraph, furnished several most efficient ministers to the
Moghal regime ; the principal supply, however, was, as it still is, from the
Kayasth caste, which, from the upper Ganges, was introduced into Gujarat
by the Muslim Viceroys and naturalised there. A similar colonisation was
begun by the same agency in the Dekkan, but the local Brahman was
there too numerous and too well-established throughout the country to
leave room for a rival, and the offshoot from the main Kayasth branch,
under the name of Prabhu, forsook the tableland for the coast, and settled
in Bombay and its vicinity. Here they were found so useful by the early
British merchants and officials that until a generation or so ago, Prabhu
and clerk were synonymous terms in those parts. In the present day the
main stron^j^hold of the Kayasth is in Lower Bengal, into which they were
introduced from U])per India. Distance, however, as usual in India, has
entirely divided the two communities, and there is no intermarriage between
the Kayasth of Bengal and his caste-fellows of Bihar and the north any
more than with those of the west coast. Even the local bodies of this
caste are much subdivided into smaller endogamous sections, generally
territorial. The position of the Kayasth and other writer castes in the
social hierarchy has long been a matter of heated controversy. In what
may be called the primary distribution of rank according to function no

Castes and Caste-Groups. A. Special Groups. 39

place could be assigned to a body which was not then recognised as
distinct from others. Literary qualifications which may well set off a
Brahman, are, by themselves, of little value as a passport to the esteem
of a public deliberately illiterate. Distinguished members of the writing
class, such as those mentioned above, were duly honoured as individuals,
but did not ennoble the community in which they were born. The dis-
proi)ortion between the ability of the writer castes and the value of their
work on the one side, and the company they were classed with in private
life on the other, grew more apparent as, under the British system of
administration, their prosperity and influence increased. It is no wonder,
therefore, that efforts have been strenuous and freciuent on their part to
establish themselves upon a social footing higher than that now recognised
by the arbiters in such matters. The line taken as that of least resistance
is the usual claim to Ksatriya lineage. There is not, however, in their
case, the probability of racial difference between them and the Indian
masses of the north and east which is lent, in the case of the Khatri and
their offshoots, by tradition, physique and locality of origin. In the parts of
the country, therefore, where Rajputs arc found in strength and Brahmanic
influence is strong, the Kayasth is a respected caste high up in the middle
classes, but nothing more. In Lower Bengal, however, where the Rajput
is a casual exotic and the weight of Brahman opinion is insufficient to
appease the jealous ferment of an inchoate social system, the Kayasth
ranks within a place or two of the Brahman, and practically, though not
avowedly, above the warrior. In Gujarat, where the clerical i)rofessions
are by no means the monopoly of the writing castes, there is, in addition
to the small colony of Kayasth, a still less numerous community called
the Brahmaksatriya, whose appearance and customs confirm their assertion
of relationship to the Khatri of the Panjab. Their immigration, indeed,
occurred as late as the 14th century. They are not only writers, but also
holders of considerable landed estates in the most prosperous parts of
the province, and their position is in many respects higher than that of
their compeers in the north. Another nominal offshoot of the writers of
the north is the Karan or Mahant of Orissa. This community is considerably
subdivided into endogamous bodies, the more southern of which retain
traces of non-Brahmanic marriage rules. It is very probable, therefore, that
tht)se nearer Bengal affiliate themselves to the Kayasth of that province,
whilst the rest remain in closer communion with the corresponding groups
of the Telugu country.

These last, with their Tamil congeners, stand on a different footing
from the writer castes of the north. The upper grades amongst them, it
is true, are strict in their observance of Brahmanic ceremonial, and wear,
occasionally at least, the sacred thread. But, like the Dravidian traders,
they appear to have arisen out of the cultivating castes, and began with
being, what most are still, the accountants of the village, a branch of
clerical work which, when not kept in the hands of Brahmans, is relegated
to the lower grade of writers or even, as in Bihar, to another caste,
and connotes an inferior social rank to that of the rest of the order.
Intermediate between the Brahman and the Karnam comes the Vidhur, of
the Maratha country, a small caste which supplements the clerical staff
of the Central Provinces and Berar. By origin the Vidhur is Brahman on
the father's side, but maternally of a lower caste. Similarly constituted
communities are found in the Konkan and other parts of the Maratha

40 5- ETHNOGRAPm'.

country. Finally, a place is found under this head for a caste difficult to
grade elsewhere, though, according to its title of Vaidya, it ought to be
dedicated to the practice of medicine. Nowadays, however, it includes
both members of other learned professions and landholders. It is only
found in Lower Bengal, where it occuiiies, thanks to the Icjcal obnubilation
of the Rajput, a position inferior only to that of the Brahman. This high
rank is due to the fact that one of the most powerful dynasties in this
part of India between the nth and 13th centuries, belonged to this caste;
and the most renowned occupant of the throne, Ballal Sen, api)ears to
have e.xercised with drastic results the regal function of making and
graduating castes, a function which in the present time is retained in
working order by the Chieftains of the Panjab Himalaya alone.

§ 31. Religious Devotees and Mendicants (-,755,900): The abdica-
tion of worldly position and the relinquishment of all possessions and
family ties, in order to jjursuc an undisturbed course of contemplation
preparatory to quitting the present existence, is a iiroceeding which has
been strongly attractive to the higher ranks of the Brahmanic community
almost from the ])ost-Vedic organisation of society upon sacerdotal lines.
Indeed, according to the strict theory of duty set forth in the treatises
dealing with the Perfect Life, it is incumbent upon every Brahman thus
to break with his former ties as he feels old age creeping over him.
Although this injunction is substantially inoperative, there are other con-
siderations which tend to swell the ranks of religious devotees in modern
India. Looking only at the lower side of the case, the vast number of
popular saints and deities, some universal, others w^ith only local renown,
is in itself an inducement to many to earn their living by invoking a
Blessing in the name of one or other of these objects of veneration upon
the households within the area of adoration, receiving in return a handful
of meal and a pinch or two of condiments. Life is easily sustained in the
tropics upon this frugal diet, whilst the climate affords opportunities for
a pleasant nomadic existence, which, if extended as it often is, to the
visitation of the chief centres of pilgrimage, brings these classes into con-
tact with their co-religionists from all parts of the country. It is no matter
for surprise, therefore, that about one in a hundred of the population has
thus taken to the road, leaving little room, accordingly, for the lay mendi-
cant, outside the ranks of the maimed, the halt, the blind and the leper.
But whilst the lower grades of the profession are laxly recruited and
the members thereof take their calling very lightly, there is in all the
principal orders a body formally initiated and put through a course of
instruction in certain tenets of doctrine and morality which they are
in turn sent forth to inculcate upon the community at large. Most of the
great orders originated in the South of India. Some are said to have been
instituted by the celebrated S'aiva reformer, S'ankaracarya. but most at-
tribute their creation to his successor, Ramanuja. On reaching upper India,
however, their constitution and practice were altered by Ramananda and
Caitanya, who mitigated to a considerable extent the exclusiveness of
their recruitment and the austerity of their regulations. The object which
these bodies were originally formed to promote was the extirpation of
Buddhism, a task begun by the great leaders of the Brahmanical revival.
Confined at first to the Brahman and Ksatriya, or Raji)ut, the orders
began, in due course, to open their ranks to members of other castes,
and then split up into two sectious, the celibate, or ascetic, and the do-

Castes and Caste-Groups. A. Special Groups. 41

mestic. The orders which admitted the lower castes too, were soon sub-
divided into the exclusive and the catholic branches, as in the case of
the Vaisnava of Bengal, jiart of whom came under the levelling influence
of Caitanya. The branch which takes to family life forms sei)arate endo-
gamous communities, and judging from the number of women returned
under the various titles, excluding certain castes which bear a name also
borne by non-ascetic bodies, such subdivisions appear to be in the ma-
jority, for there are in the aggregate 90 women to every lOO men. In
Bengal, indeed, the former are in excess, as they are in the population
at large in that province. In upper India, however, there are many large
establishments of the nature of monasteries which supply the bulk of the
higher grades of itinerant teachers. Even in these, however, the functions
of the fraternity are not restricted to religion, for some of the Wahantas,
or Abbots, as they have been called, have been noted money-lenders on
the strength of the funds and endowments of their charge. In former days,
too, bodies of these devotees used to be formed into irregular forces,
which exhibited in action the same fanatical ferocity as is now associated
with the Muslim Ghazi and in the middle of last century with the Sikh
Akali. A remnant of one of these bands still survives, it is said, in the
Dadupanthi Naga of the State of Jaipur in Rajputana, a country associated
to some extent with the expansion of the ascetic movement. It is not pro-
posed to enter here into the doctrinal differences between the various
fraternities further than to mention that there is the usual main division
of the principal bodies into S'aiva and Vaisnava, with many subdivisions,
the latter school being the more modern. Nor, again, is it necessary' to
set forth in detail the sections of the orders, since being recruited from
all classes of the population, regardless of caste or race, they are of no
ethnographic importance, and under each head are included members of
the Sikh, Jain and Muslim creeds along with those of orthodox Brahmanism.
It is impossible, indeed, to state accurately the numbers falling under
each head, owing to the loose way in which the principal designations
are applied. Under the title of Fakir, for instance, which is specially ap-
plicable to Muslim devotees, nearly 450,000 Brahmanists and Sikhs are
returned. The Atit, again, a general title, are given as identical with
Gosavl or Sannyasi as well as under their own heading. Vairagi or BairagI
covers not only the Vaisnava and some of the Dandasi, but also most of
those returning themselves as Bhava or Sadhu, terms used of Brahmanic
devotees in general. Still more misleading is the return under Jogi, an
order differing from the rest in its origin, and conjecturally not called
into existence to combat schism, but itself a heretical order, proscribed by
the orthodox, probably on account of its then Jain or Buddhist proclivities.
It is shown in combination with the Jugi, a class of coarse-cotton weavers
in eastern Bengal and Assam, reputed to have come from the south-west,
but undoubtedly taking its rise from some religious organisation of the
lower classes, and now said to be "assuming the sacred thread en masse",
and contesting its right to wear it against the local Brahmanity. In upper
India, the Jogi or Yogi community is divided into those who have a right
to the title by profession and initiation and others who have assumed it
for the convenience of their calling. The former, of whom there are two
main subdivisions, have their monasteries and settled organisation, the
latter who are returned in the Panjab, Rajputana and Gujarat under the
name also of Raval, trade upon the reputation the other Jogi have acquired

42 5- Ethnography.

for obtaining supernatural powers of divination by dint of contemplation
and mental abstraction; consequently, "any rascally beggar who pretends
to be able to tell fortunes or to practice astrological or necromantic arts
in however small a degree, buys a drum and calls himself a Jogi". The
43,000 Muslim returned as Jogi in the Panjab and its neighbourhood arc
thus accounted for. C(jnsidering the Dravidian origin of most f>f the ascetic
orders and the traces of the South still preserved in their customs and
nomenclature, it is remarkable that hardly any are now found in that part
of India, and those chiefly of the lower class. Even the mendicants who
there ply their trade in the name of religion hold no reputable position
in the community. This is perhaps attributable to the fact that though the
genesis of the great orders took place in the south, it was in the north
that the need of their propagandist efforts was most pressing.

B. The village community.

§ 32. In the greater part of India, the village as a unit not only
of population but of land, has assumed a form not to be found in other
countries. In European Russia, it is true, the system of rural aggregation
bears a considerable resemblance to that of India, but has far less weight
in the social organisation, and is far less bound up with the ethnic evo-
lution of the country. The village, then, as it falls within the scope of this
review, is an agricultural community on a self-sufficing basis, congregated,
for the original purpose of protection, on to a single site, surrounded by
a definite area of land the prescriptive right to which is invested in it.
Originally, no doubt, the occupants of the soil formed a close corporation
based upon kinship or common descent, but in the course of time that
exclusiveness crumbled away, and new comers were admitted to the land,
though on an inferior footing, in most cases, to the rest. The village exists
for the agriculturist, and the exercise of other callings therein depends
upon their necessity or utility to him, and this, in turn, depends upon the
relative isolation of the village from other sources of supply. The staple
staff of artisans and menials is remunerated directly from the soil in re-
cognised proportions of the harvest, so much threshed grain from each
landholder. The completeness of the organisation varies considerably in
different parts of the country, but where it exists, its main features are
much the same. The village, in the above sense, is not found in the
comparatively recent settlements east of Bihar, or on the Malabar coast;
nor has it taken root amongst the more or less migratory tribes of forest
tracts, where the insufficiency of arable land and the frequent flittings of
the population from spirit-haunted or unlucky locations are adverse to so
stationary an institution. Although, then, these tribes live mostly by rough
methods of tillage, they cannot be counted amongst the landed classes,
and arc therefore dealt with apart from those to whom that designation
is conventionally more appropriate. The latter can best be considered
under two heads, first, the castes which hold their land as a military or
formerly dominant body, and, then, the peasantry dwelling alongside of
them without traditions of a status or calling other than that which they
now enjoy.

§ 33. Landholders, Military or Dominant (23,702,400): Castes of
this type may be e.^cpected to be more powerful and more prominently
demarcated from the rest in the track of the great racial inroads from

Castes and Caste-Groups. B. The Village community. 43

the north-west. Eastwards of the settlements of the Pathan and Baluch
tribes, which will be treated of in a later paragraph, a line drawn from
the Gujarat peninsulas, through Malva, to the Ganges, marks off the domain
of the Rajput, Jat and kindred tribes, whilst the Salt range of the Panjab,
and the sub-Himalayan tracts from the Jehlam to Nepal form their general
limit on the north. East of Bihar, the Koch, or Rajbansi, and the .AhOm
may be said to occupy a somewhat similar position amongst the Mongo-
loidic population. In the Dekkan, the Maratha may be included on histo-
rical grounds in this category, though his origin is doubtful and the limits
of his caste wanting in definition. In Southern India the title of dominant
is applicable to several Dravidian communities which rose into prominence
with the dynasties of which they constituted the chief military forces, and
on disbandment, either reverted to or assumed, the position of cultivators.
There is no question here, therefore, of foreign origin. On the Malabar
coast, on the other hand, the Nayar, though now thoroughly Dravidianised,
is said to have come from the north. In Orissa, again, the Khandait makes
the same claim, but is probably of much the same origin as the other
Dravidian communities of this class.

Reverting to the castes of ujiper India, the Rajput has been dealt
with in a former paragraph sufficiently for the purposes of this review.
An important point in connection with the subject immediately in hand
is the close connection between the Rajput and the Jat, who ranks next
to him both in numbers and position throughout the Panjab jilains, Raj-
putana, and the upper Ganges and Jamna valleys. It has been conjectured
by some that the difference between the two communities is social, not
racial, the Rajput being a Jat leader who, after being successful in the
field or on his estate, bound himself and his family to the strict obser-
vance of Brahmanic rules and thus attained the pinnacle of orthodo.x
repute, whilst the rest of his tribe remained Jat in name and in their
traditions and practice. In the circumstances of the two castes in the
Panjab in the present day there is much to support this view. Others hold
that the Jat belongs to a later wave of immigration than the Rajput, and
entered the Panjab from the west, by way of Sindh and the Indus, whilst
the Rajputs were still in Rajputana and its eastern neighbourhood. However
this may be, the northern stock has now been fused, and though the Jat
no longer becomes a Rajput, the same tribe is found Rajput in one village
and Jat in the next. In the Jamna tracts this is not the case. Whether
because the Jat arrived there direct from Sindh and remained at a distance
from the seat of the predominant body of his tribe, or whether by reason
of admixture with inferior Rajput blood, his physique and social position
are lower. The Jat par excellence is the peasantry of the Sikh tracts,
where the tradition of political supremacy is still green, and the Jat has
nothing to gain in public estimation from either Brahman, Rajput or Pathan.
Along the Jamna, he has succumbed to the prevailing influences, and looks
up to the Rajput, whilst in the west, he does the same to the leaders of
Muslim society, and his name has been there bestowed upon any cultivator
of that religion, whatever his caste. Like the Rajput and other great com-
munities in the north-west, the Jat places religious considerations beneath
tribal in his domestic arrangements, so it appears from the Census that
one third of the population bearing this name are Muslim, one fifth Sikh,
and just under half, Brahmanist. As stated above, the Jat is in the first
place a cultivator, and the women of his family share to the full his

44 5- Ethnography.

enthusiasm in the pursuit of the family calling. The Sikh Jat is also a
born soldier, not merely a combatant, but a disciplinarian, and equally
efficient on the snow-clad ridges of Afghanistan and the steamy plains of
Tientsin. Next to the Jat in rank, and probably akin in origin, comes the
Gujar, a caste as to whose descent there has been much controversy
between the pro-Aryan and the [iro-Scythian. The caste is now generally
affiliated to the Gurjara, a tribe which was settled in the neighbourhood
of the Caspian, and entered India either in company with or at the same
time as, the Yetha or White Huna, of whom they arc said to have been
a branch. They spread very widely over the west and north-west, and one
body of Gurjara obtained a dominant footing in the'western province which
is now called after them. Their connection with it, however, after the
downfall of their dynasties, was dissipated into innumerable channels of
castes, where it is recognisable only in customs and in the titles of some
of the sub-castes. The greater jiortion of the Gujar settled in the Panjab
and along the Jamna, with a considerable colony in Oudh. In the first
named tract, again, they have left their name behind them in several
places, but it is only in the submontane portion that they can now be
called a dominant tribe. In the plains they follow their traditional occu-
pation of cattle-breeding, combined, it may be, with cultivation, in which
they are not so expert. Their unrestrained devotion to the horned beast
is such that in some parts of India their title is derived from the Sans-
kritic term for Cowthief. Even though philology may not support this deri-
vation, it has the authority of their almost universal reputation. They are
not now found south of the Vindhya, where those returned as Gujar are
traders from Gujarat, who, as stated above, retain traditions of a cognate
origin. It is held, indeed, that a Gurjara element underlies all the chief
cultivating classes of Gujarat above those traceable to a distinctly K61
origin. Returning to the Panjab, the south of the Salt range tract is the
present home of the Avan, who have been there for at least 6oo years.
They are said to have come up from Marvad or upper Sindh, and to have
belonged to one of the numerous Scythic bands which gave the Jat and
other castes to the country further east. Though the Avan are nearly all
Muslim, they retain Brahmanic names in their genealogies, and use Brahmans
as their family priests. They have not spread beyond the north-west corner
of the Panjab, where they share with the Janjhua Rajput and the Khokhar
the predominant position among the peasantry. The Khokhar, however,
though equally of the faith of Islam, have maintained more fully the
tradition of Rajput origin, and return themselves in considerable numbers
as a clan of that great caste. Others, again, claim to be Jat. The Gakkhar
in the north of the Salt Range plateau are similarly situated to the Avan
in the south. There seems to be little doubt but that the three tribes are
all of allied Scythic origin, and became Rajput during the Brahmanic
revival, Jat when the Sikhs rose to power, and claimants to Mughal blood
now that the influence of Islam reigns supreme in this region. Among the
tribes behmging traditionally to this part of India may be counted the
Kathi, though in the present day they are found under this title only in
the western peninsula to which they have given their name, and even
there in but small numbers. In the Panjab they consider themselves a
subclan of the Panvar Rajput, and are thus merged in the general mass
of that order. In Kathiawad they preserve the tradition of migration from
BFkaner and Multan, the latter being the very tract in which they were

Castes and Caste-Groups. B. The Village community. 45

found by Alexander as a foreign nomadic body, successfully resisting the
expeditions sent against them by neighbouring Aryan potentates. It is
conjectured, therefore, that these, too, are Scythic tribes connected with
the rulers of Taxila at that period, and were driven into exile through
Sindh into Kach by the Muslim invasions. They are now principally
cultivators, but keep green the remembrance of their original occupation
of cowherds by breeding horses and cattle. They also retain their ancestral
sun-worship, and a rude representation of that luminary is affixed to all
their formal documents. It is not improbable that they are of the same
stock as the Ahir or Abhlra, the great cattle-breeders of upper India,
though their position is now higher than that of the latter. In Sindh, two
Rajput tribes of agriculturists, the Sumro and the Sammo, successively
occupied the dominant position on the lower Indus from about 750 A. D. to
the middle of the i6th century, and now belong to Islam. Their respective
numbers are by no means accurately represented in the Census return
owing to the wide-spread practice in this province of giving the general
title of Sindhi as the name 01 the tribe or caste, thus placing nearly a
quarter of a million of the inhabitants beyond the possibility of identification.
East of the Panjab, the only caste, beside those already mentioned,
which can be described as dominant, is the Taga, a community of the
upper Jamna. Its origin is doubtful; though it seems to be generally agreed
that it has Brahman blood; but the prominence of snake-worship amongst
Taga, together with the division of the caste into the "Score" and Half-
Score" sections, indicates considerable admixture of local races. Their
degradation from Brahmanical rank is attributed to their addiction to
agriculture, as in the case of the Babhan of the south-east. More than a
third of them are now Muslim. In Bihar, the only dominant caste beyond
the Rajput is the Babhan or Bhumhar, already mentioned in connection with
Brahmans, which forms but a small proportion of the population. Lower
Bengal as above stated, was never colonised by military occupation, and
the only caste which may be called dominant is the Koch of the northern
territory bordering upon the Brahmaputra. Their claim to this position
rests upon the long existence of the Koch kingdom of Kamarupa, in the
Assam valley, and its extension, for a time, into Bengal. The latter portion
was separated from the rest towards the end of the i6th century, and
succumbed to the Muslim, as did the other shortly afterwards to the Ahom.
There are two distinct sections of the population owning to the name of
Koch. West and south of the Brahmaputra it is said to be of Kol-Kher-
vari origin, and has long been Brahmanised under the designation of
Rajbansi, which satisfies the aspiration of the local peasantry, as that of
Rajput crowns the ambition of the Chieftain or large landowner in other
parts of India. In Assam, on the contrary, where the lineage of the local
leading families is known, the Koch is IMongoloidic, or Bodo in origin,
and its rank and file are recruited from all the Bodo and iSIikir tribes of
the valley, who drop their own title on adopting Brahmanism. Some go
further, and pass at once into Rajbansi, or embrace Islam if their claim
be not allowed. The respective numbers of the two are, 2,115,700 Rajbansi,
chiefly in Bengal, and 292,100 Koch, of the Assam branch. The Ahom of
the more eastern portion of the Assam valley, are also a once dominant
tribe of agriculturists of Indo-Chinese descent, who will be referred to
under the head of Assam Hill tribes. There is one more caste belonging
to Bengal which may be here mentioned, to wit the Khandait of Orissa.

46 5- Ethnography.

They seem to have been originally a body of local militia enlisted from
the Bhuiya, a Kol tribe, and commanded, probably, by officers imported
from upper India. Some of the customs of the latter commended themselves
to their subordinates, on the strength of which form of flattery, a claim
to the caste of Rajput was subsequently advanced. The Khandait is divided
into two sub-castes, one comprising the landholders, probably endowed
with estates for military services ; the other the peasantry and village
watchmen. The former hold a good position and rank next to and but
little below the Rajputs, who, as elsewhere in Bengal, have not taken firm
root in the soil. A community which once carried its arms not only into
Orissa but up to the very walls of Calcutta, without leaving any enduring
trace of its passage, is the Maratha, the principal landed class in the
Dckkan, and the dominant power in Baroda, Gwalior, and practically in
Indore and several other states. The origin of the Maratha is obscure.
Elsewhere in this work it has been stated that recent anthropometrical
observations have given rise to the conjecture that there is a Scythic
element in the population of the Dekkan beyond that which can be attri-
buted to the dynastic influence of the various Ksatrapa Chieftains who
maintained their power there long after the dissolution of the Huna
sovereignty in Central India. The Brahmans of upper India, too, have the
belief that the Maratha arc of Persian descent, and that the Citpavan
Brahmans of the Konkan were their sun-priests, introduced in the 7th
century and formally adopted into the local hierarchy. However this may
be, there was not imjirobably some distinction between the masses and the
dominant classes based upon race, as in Rajputana; but it did not obtain
prominence until the leading families were welded into a military body by
the Bhonsla. S'ivaji donned the sacred cord and took the title of Ksatriya
upon his enthronement, and within a generation, his successors made a
claim to definite Rajput descent, and were apparently not rebuffed even
by the highest of the Rajput Chieftains. The kinship, however, has not
been practically acknowledged, possibly because the political atmosphere
has changed since the beginning of the iSth century. In the present day
there is no definite line drawn between the Maratha and the Kunbl, or
cultivating peasantry-, though the leading clans of the former still enjoy
special consideration. Recruitment admittedly takes place from below, and
any KunbT who prospers above his neighbours, renounces widow-
marriage, secludes the women of his family, marries his daughters at an
early age and within a narrow circle, and puts on the sacred thread for
special occasions, becomes in due course a Maratha in title, with hyper-
gamous tendencies not always ignored by the older families. Both Maratha
and Kunbl are distinguished by the totemistic, not Brahmanic, character
of their exogamous subdivisions, and by their worship of the same local
deities, so that, like the Jat, the upper classes may have assumed a
distinct po.sition without imposing the impassable barrier which exists
in the north between the Rajput and the rest. Amongst the Marathas as
a whole the only barrier of that nature is geographical, a Dekkani not
intermarrying with a family in the Konkan, in spite of the identity of
language. The climate, which entails a difterence of cultivation and con-
sequently of diet, has affected the physique, and the broad-acred grower
of millet disowns the tiller of the petty rice-patch.

§ 34. The Dravidian country remains to be considered. In the greater
part of this tract the military and dominant element in the landed classes

Castes and Caste-Groups. B. The Village community. 47

is insignificant. In the Karnatic, indeed, it is scarcely to be found, and
in Telingana, too, the position of a special subdivision is often found to
rest upon the military recruitment of a former dynasty. The Razu, who
were settled in the extreme south of the Tclugu country by the Vijaya-
nagar C'hiefs, for whom they fought, seem to have the best claim to the
distinction in question. They are undoubtedly superior to their neighbours
in physique, and are more scrupulous as to ceremonial. They wear the
sacred thread, seclude their women and employ Brahmans as their family
priests. It is not improbable, therefore, that they are the remnants of a
body of mercenaries from further north, and really differ in race from
the Dravidians with whom they are now permanently associated. The
Velama of the north Coromandel coast are an offshoot of the great Kapu
or Reddi caste and closely connected with other agricultural bodies of
the neighbourhood. They have amongst them, however, several wealthy
and influential Zamlndars, or landed proprietors, and having adopted Brah-
manical regulations more strictly than the rest, are generally considered
to hold a somewhat higher position. In the Tamil country, especially in
the south, the line of demarcation between the military castes and the
others is more easily traced than amongst the Telugu masses, but there
seems to be this noteworthy difference between the two regions, that the
immigrant peasantry of the south rank higher in the present day than
the castes once dominant, so that, setting aside the Chieftains and Zamin-
dars, there is the tendency for a landowner of the latter, as he advances
in prosperity, to get merged in the ranks of the former. The popular
version of this inclination runs: "The Kalian became a Maravan ; the
Maravan became an Agamudaiyan, and the Agamudaiyan is now a Vellalan".
The explanation seems to be that the formerly dominant classes obtained
their position by predatory, rather than military, prowess under the weak
governments of the past, and retained with their independence their
original religion and customs. In the piping times of the j^ax Britannica,
however, Brahmanic influence is permeating the masses, and as its cere-
monial is the touchstone of respectability, the more aspiring remnants of
the earlier civilisation affiliate themselves to a body already in full touch
with the refinement aimed at, in preference to taking up the invidious
position of innovator in the community of their birth. The principal tribe
coming under this head is the Kalian, which happens to be the Tamil for
thief It is probable that the original meaning was different, but no alter-
native has been found, and the interpretation is unfortunately justified by
the history and habits of the caste. It is conjectured that the Kalian are
an offshoot of the great Kurumban, or cowherd race of the south, which
spread downwards from the uplands of Mysore, and were ousted from the
plains successively by the Cera and the Cola dynasties. Some of the tribe
expelled in their turn, the peasantry introduced by the latter, and settled
on their lands. The reputation thus acquired helped to keep the Kalian
in independence, and enabled them to maintain to this day their old customs
untainted by Brahmanism in their essential features. The acknowledged
head of their tribe is the Raja of Pudukottai, called by them the Tondaman,
in memory of their former colonisation of Tondamandalam or the Pallava
country. The bulk of them are cultivators and labourers; but they still
furnish a strong contingent of watchmen, a duty which serves them as
the pretext for the levy of a prophylactic subsidy from the householders
thus subjected to their protection. Their neighbours to the south, the

48 5- Ethnography.

Maravan, arc amongst the earliest inhabitants of this tract, and at one
time got possession of the whole of the Pandya or Madura domain. They
furnished a strong body of militia, and for many generations lorded it
over the rest of the population. There is some connection, at present un-
ascertained, between them and the Kaljan. Like the latter they worship
their own gods and demons, and employ for the purpose priests drawn
from the lower castes, but for ceremonial other than that of the temple,
they call in Brahmans. Their head is the Zamlndar of Ramnad, to whom
the Tondaman and other local magnates do obeisance when they meet.
The Agamudaiyan again, are closely connected with the Maravan, with
whom they intermarry under rules which in the Brahmanic system would
imply hypergamy in favour of the latter. Nevertheless, the Agamudaiyan
is the only caste of the three which has been substantially Brahmanised,
and in many ways it comes near the Vellalan in practices and beliefs.

Crossing the Peninsula, a distinctly dominant class is found in the
Nayar of the Malabar coast, a community of northern race, with uncertain
traditions as to its original home or the route by which it reached its
present secluded domicile. It has its own peculiar customs and institutions,
which, as in the case of the Rajputs, have been assimilated by indigenous
castes of lower rank, who thereby justify the arrogation to themselves of
the title of their superiors. The community, therefore, no longer consists
of military landowners, as formerly, but includes, under subdivisional names,
not only artisans and traders, but even menial castes such as the barber
and washerman, who have found it worth while to devote their services
exclusively to the Nayar. It is probable, then, that not more than three
fourths of those returned under the latter title are true Nayar, and that
these belong to at most three subdivisions of the tribe. The customs of
the Nayar are, as observed above, peculiar, and of high ethnological in-
terest, but it is not within the scope of this review to enter into them.
It may be remarked in passing, however, that in many of them may be
found traces of polyandry. Inheritance is through the female. The exogamous
unit is based on descent from a common female ancestor in that line.
The endogamous limit is hypergamous for the female, and either within
or below the subcaste for the male. The Nayar of the north and those
of the south form separate communities, the division being evidently based
upon the notion that pollution lies in the south, perhaps because that
region is further from the caste-cradle. The distinction between the tAVO
is so strictly enforced that though Nayar males may circulate freely over
the whole country, no female of the northern section may cross the river
which divides Kanara from Malabar, nor, again, that which intersects the
latter district. This group is completed by the addition of the Kodagu,
or dominant tribe of the little district of Coorg, not by reason of its
numbers, but, like the Kathi, because it has had a history, and has managed
to maintair. its ])osition and language in its native uplands against all
comers. Since the tract has been opened up by European enterprise, for
the growth of special products, there has been a considerable influx of
labour from Mysore and the coast, and the Kodagu now constitutes but
a fourth of the population; but that fraction is at the top.

§ 35. Peasants (36,251,100): In nearly every part of India this group
is the largest, and, together with those of the landless labourer and the
village menials, includes the bulk of the rural population. The exceptional
tracts are Rajputana and the Panjab, in which, as pointed out in the

Castes and Caste-Groups. B. The Village community. 49

preceding paragraph, the military tribes have retained their grip on the
land. In accordance with the general scheme of exposition, it is proposed
to subdivide this group here into the cultivating castes, in the wider
sense of the term, and those who devote their efforts to the growth of
special products, such as the bitcl-vine, or to roots and vegetables and
other market-garden produce. Of the latter some are conventionally im-
pure, such as onions, turmeric and turnips, or necessitate the destruction
of life or extensive and intimate dealings with manure, both rejiugnant
to Brahmanical tradition, in consideration of which the castes thus engaged
have been relegated to a lower social position than the field operator.
In the Panjab castes of this class are numerous, and in the plains
of that Province there are but two others, outside the ranks of the do-
minant, which call for mention here. The Kambo, one of the most
skilful cultivators of the province, is found along the .Satlaj and in the
east, where he has crossed over the Jamna into Rohilkhand. The caste is
of local or Kashmiri origin, though the INIuslim minority in it claim to be
Mughal. It is probably connected with the great gardening caste of the
Arain, but its position is higher. One of its sections has taken to trade
and the clerical professions, in which, however, they are said to be more
skilful than honest. The Meo, or Mcvati, is the dominant caste of a jjortion
of eastern Rajputana and a small tract in the south Panjab. It is no doubt
a branch of the forest tribe of the Mina, but having become Muslim and
acquired land, it has set up for itself. Formerly it gave much trouble from
its unruly habits, but since its larger settlements were broken up into
detached villages it has sobered down. Islam sits very lightly upon the
IMeo, and he observes the Brahmanic festivals impartially with those of
his own creed, ignoring the fasts of both. He continues to worship his
old village gods and to employ Brahmans as his priests, but in these
respects he does not differ from the bulk of his fellow converts in the
neighbourhood. In the sub-Himalayan parts of the Panjab and the outer
ranges there are a few interesting agricultural tribes on the borderland
never occupied by the Jat and the hill country of the Rajputs, never oc-
cupied by the Muslim. Some of these, the Thakar, Rathi and Raut,
are undoubtedly related to some of the Rajput clans on the one side, but
are merged into the lower Hill tribes, on the other. It is open to question,
for instance, whether the Thakar is a low Rajput or a high Rathi, and
whether the latter is not a somewhat elevated Kanait. The Raut, who is
located nearer the plains that the rest, occupies a lower rank, and though
recognised as a connection of the Candel Rajput, is more often associated
with the Kanait. The latter and the Ghirath are the chief cultivating
classes of these hills. The Ghirath is found principally in the Kangra
valley, and is noted for growing rice wherever the land is sufficiently
depressed to allow of the collection of sufficient water for the purpose.
The caste is so subdivided that the saying goes that there are 360 sorts
of rice and the same number of Ghirath clans. They are inferior in physique
and mode of life to the cultivators of the higher valleys, and though they
may have a tinge of Rajput blood, imparted by refugees from the plains,
they are mainly of the specific hill type which prevails from the Indus
to Sikkim. The Kanait are a more distinctive community of this race, and
whilst one of their two main subdivisions has become more Brahmanised
than the other, and pretends to be the progeny of Rajputs by Hill women,
there seems reason to think that they belong to a very early wave of

IndoAryan Research. II. o. 4

50 5- Ethnography.

northern immigration, possibly Aryan, but not of the Vedic branch, which
has received an infusion of other northern blood since its settlement
in the Himalaya. They are now the tenants and labourers of the Rajput
landowners. Further to the east, however, their relatives, the Khasiya of
Kumaun and Garhval, escaped Rajput overlordship, and themselves sub-
dued a lower and more primitive tribe, probably the pom. Owing to the
fact that their territory contains the two celebrated shrines of Kcdarnath
and Badarinath, at the reputed sources of the Ganges, the Khasiya have
long been thoroughly Brahmanised, though the transition from a lower
to a higher grade is more easily achieved than in the plains, and is
here the result of the acquisition of wealth, not| as in the Panjab Hills,
of royal favour. The Khasiya do not figure separately in the returns, as
they are all included under the general head of Rajput, but their number is
not far short of half a million. The community which goes by a somewhat
similar name in Nepal is distinct, and of admittedly mixed origin, Brahmanic
and Mongiiloidic Himalayan.

In the Gangetic Doab, Oudh and Bihar, the great peasant castes are
more or less connected with each other by origin, but in so fertile a tract,
well provided with large towns, the occupation of market gardening has
diverted an unusually large number of subdivisions from field work. Of
those who have clung to the elder branch of the profession, the Kurmi
is the most widely sjjrcad, especially along the Ganges and to the south
thereof. The title corresponds to that of Kunbi, used in the Dekkan and
western India. The derivation is uncertain, and though the word is found
in the form of Kutumbika in some early inscriptions, this is probably only
the Sanskritiscd version of some older name, such as that of Kul, a
Dravidian name for a cultivating landholder, in which sense it is still used,
and not only in the Dravidian country. The Kurmi is by no means a
homogeneous body, and is not only much subdivided in the tracts where
it is ajiparently of one race, but is used on the borders of the Central
Belt as a sort of occupational title for those of the Kol tribes who have
been long settled as cultivators and have thereby thriven beyond their
ancestors. Closely allied with the Kurmi by origin, though now entirely
distinct, are the Koeri. They rank below the former, who will drink, but
not eat or intermarry with them, possibly because the Kocri have succumbed
to the lucrative attractions of special cultivation, such as that of tobacco,
the poppy and even vegetables. The Kisan, again, belong to the same
slock, but like the Koorl, have long been formed into a separate caste,
and are even more exclusive in their intercourse with outsiders. There
is another community of the same name, though sometimes called Nagcsia,
who have been combined with these in the Census return. They inhabit
parts of Chutia Nagpur and the Central Provinces, and are of the Kol race.
The Lodha is a caste of inferior position and probably of earlier settle-
ment than the KurmT, from whom it differs in both physique and habits.
The Lodha are specially addicted to the cultivation of rice, and are found
nearly all over the Upper Provinces and a little way into Bihar. But the
section which inhabits BundOlkhand and its neighbourhood is probably
nearer the original stock, assuming the latter to belong to the Central
Belt, and takes a lower place in society accordingly. The cultivating
classes of the Central Provinces are those of the Dekkan in the west, and
of the south Ganges-valley in the north, with a large substratum of the
more civilised forest tribes in most parts. In the Chatiisgarh districts, the

Castes and Caste-Groups. B. The Village community. 51

Kavar is probably an offshoot of the last named group, though the
fertility of the country has enabled it materially to improve its position.
This caste, as well as the Kirar, claims Rajput origin, and there is some
ground for believing that the tribal ancestors belonged to some military clan
which settled in the hills, and thus lagged behind the rest in Hrahmani-
sation. Tiie Kirar are admitted to be Rajputs of a low class in the Jamna
valley, but are repudiated by the Rajputs of Central India and the Narbada
valley. On the Orissa border, the Kolta are in occupation of the best
lands and prosper accordingly. As they found it necessary to spread, their
keen scent ft>r the best settlements brought them into conflict with the
wilder tribes, but they held on to what they had got. In the Assam Valley,
as in the Central Provinces, the f<iundatitm of the jiopulation is a more
or less Brahmanised community of the local stock, Kol-Dravidian in the
one case, Mongoloidic in the other. In the preceding paragraph the Koch
has been mentioned as the prevailing caste in the western jiortion of the
old Kamarupa territory. Less numerous but of higher positiim in the
present day are the Kalita, an immigrant caste, or more correctly, tribe,
for they jirobably entered the valley before the caste system had been
fully developed in Bengal. Though the Kalita are mainly husbandmen,
they do not constitute a caste in the strict sense of the term, (or they
exercise all the crafts and occupations which are elsewhere relegated to
endogamous functional bodies. The usual tendency to specialise, however,
is not absent, and subdivisions are being formed upon the normal lines.
Kalita, too, is becoming, like Koch, a designation of social rank, and
lower communities are assuming it, either by absorption or as distinct
units. Outside the ranks of the forest tribes, the only other agricultural
community which need be mentioned here, is the Halvai-Das, of the
southern or Bengal valleys. This, in Bengal, is accounted a subcaste of
the great Kaibartta community, but in Sylhct, and in such parts of the
Brahmaputra valley as it has reached, it has succeeded where in Bengal
it failed, in establishing itself as a separate caste of higher position
than the body from which it rose. Its prosperity has brought it, as is not
uncommon, a superior marriage field, and girls of the Kayasth and
Vaidya castes are given, albeit under protest from outside, to well-to-do
Halvai-Das. Ne.xt generation will possibly see a still further advance
sanctioned by the somewhat fluid public opinion of the two Provinces
concerned.

The enormous population of Bengal furnishes, as is to be expected,
a good number of large cultivating bodies, many of them, as was above
pointed out, nourishing claims and aspirations which would be futile in
an older Brahmanic society. The most populous of all, the Kaibartta,
accounts its agricultural sections far above those which fish, and has framed
its subdivisions accordingly. It is doubtful which occupation is the earlier
amongst them, but from their appearance, it is surmised that they are
immigrants who spread over the Delta, from the country round Midnapur
and took to fishing for a livelihood as their numbers increased. St>me of
the larger landed proprietors are said to have become Rajputs. In Orissa,
some became Khandaits, whilst the Casa, one of the i)rincipal sections,
has invented the name Mahisya for itself, to which its claim has been
acrimoniously disputed. The Sadgop is most numerously represented in
and about the same tract as that which the Kaibartta regard as their
early home. It is supposed to have abandoned cowherding, as the Kaibartta

52 5 Ethnography.

abandoned fishing, in favour of agriculture. The more prosperous Sadgop
are said to be dropping the plough and employing labour on their land,
thus paving the way for a higher endogamous subdivision. The caste
stands higher in rank than the Kaibartta, owing probably to the superior
purity of their traditional occupation. Like other Bengal agriculturists,
they are sometimes called Casa, a general term, like that of Kurmi or
Kunbi. There is, however, a caste in Orissa to which the name of Casa
is specially applied. It is of K61 or Dravidian origin, and whilst admitting
members of other castes to its lower ranks, passes in the u])per into that
of Karan or Mahant, mentioned above as the lopal writer caste, on the
way to establishing touch with the Kayasth. The Gahgautais a small but
respectable caste of north Bihar, much the same in position as the Kurmi,
but ranking below them, and more lax in their diet. Round Calcutta is
found the fishing and cultivating caste of the Pod, lower than those above
mentioned. Like the rest, however, it has its lower and ujjper endogamous
subdivisions, the latter of which pul in their claim to Ksatriya lineage. Most
of the caste are cultivators, but some have acquired considerable estates,
whilst others have taken to trade and handicrafts. It appears to be con-
sidered to be of Deltaic origin, like the Candal, as the Brahmans who
minister to it are avoided by their fellows, but those who only act as
teachers remain unpolluted. The Candal or Namasudra, is the largest caste
in eastern Bengal, and, as its name suggests, stands very low in the social
scale. It is much subdivided, and eight of its main subdivisions are func-
tional, and never eat and seldom intermarry with each other. The agri-
cultural section stands out from the rest in rank, and next to it comes
the boating division. Fishing, however, except for the domestic larder, is
strictly prohibited. The Namasudra employ a special class of degraded
Brahman of its own, and its barbers and washermen are also members of
the caste. The Census was made the occasion of an attempted severance
of the caste into S'udra, the superior body, and Nama, the Bengali for
"low", to include the rest. It failed.

§ 36. In the Dekkan and adjoining tracts, the one great cultivating
caste is the Kunbi, which has been already treated of in connection with
the Maratha. Like every caste spread over a wide area it is much sub-
divided, but its position and general constitution are fairly uniform. The
corresponding caste in Gujarat, which has been included under the general
title, calls itself Kanbi, and is distinct from the Dekkani in origin, and
custom as in language. Along with the tradition of early immigration
from the north, it has many points of resemblance with the Gujar of the
Panjab. The Kanbi is almost entirely agricultural, and is in occupation
of the most fertile tracts of Gujarat, with the reputation of making the
most of them. The only alternative occupation generally recognised is
silk-weaving, to which one of the subdivisions is devoted. A branch of
the Kanb'i is settled in the north Dekkan, an ancient domain of the Ahir,
or cattie-brceders. Here the caste is known by its old name of Gujar,
but its subdivisions are those of the modern caste of the coast. The
Khadva Kanbi, one of the main subdivisions, has the custom locally pecu-
liar to itself and the Bharvad shepherd, of celebrating its marriages only
once every ten or eleven years, according to the vaticinations of their
chief sacerdotal advisers. Naturally, so rare an opportunity has to be seized
irrespective of the ages of the children, so that not only are infants in
arms duly betrothed, but women in the family way join in perambulating

Castes and Caste-Groups. B. The Village community. 53

the nuptial altar, on the understanding that their future offspring, if sex
permit, arc thereby made man and wife. As to the relative number of the
I\Iarathi and the Gujarati sections of this caste, there are probably at^out
2,700,000 of the former, and 1,350,000 of the latter. The Koll in its
various subdivisions is probably an early dark race extruded from the
plains by the Kanbi, so far as it is found in the west. Under the same
name, it is found from the Panjab Himalaya to the Sahyadri Ghats, not
to mention the Kol of the Central Belt. In the first named tract Koli is
a general term for the menial classes, amongst whom most of the artisans
are included, in Gujarat there is a coast Koli, generally a boatman or
fisher, and a large landed class, chiefly in the north of the province, called
either Talabda, the Locals, or Dharala, the arm-bearers. Some of its clans
intermarry with the lower Rajputs, whose rules of exogamy they have
adopted. In or near the hill country the Koli approximates to the Bhll,
though perhaps more settled in habit. On the Sahyadri, however, their
reputation is lower, and the Malhari branch are apparently the descendants
of a wild tribe of the south western Belt, driven westwards by the ad-
vancing Muslim or by colonists from Telingana.

In the Dravidian country, the castes are remarkably well demarcated
by the linguistic divisions, and whilst there are considerable colonies of
the northerners in the Tamil country, the reverse movement appears to
have been ver)- trifling. In the Karnatic tableland, the cultivating castes are
found under a few general headings, such as Vakkaliga and Lirigayat,
the former in Mysore, the others further north and east. Under the Lingayat
or Lingvant, system, caste is supposed to be merged in the general title,
and though this rule was followed to a great extent at the Census, in
practice, caste is recognised almost as fully as amongst the orthodox
Brahmanists. The community, as a whole, falls under three heads; the
original converts of Basava, with a few later additions ; the functional
group, and, lastly, the impure castes of village menials. Each section has
an amazing number of subdivisions, since nearly every one of the local
Brahmanic castes has its Lifigvant subdivision, endogamous and distinct.
The general tendency in the present day is to assimilate the Brahmanic
organisation under the Jangam, though occasionally the upper classes in-
troduce Brahmans as priests. There has been a movement, indeed, to get
the whole community recorded as VIrs'aiv, subdivided into the mythical
four Varna of the Purusa-Sukta. Irrespective of the latter refinement,
the first suggestion refers to a time anterior to the founder of the
sect, and in supersession of the usage of centuries. There are a few
Lingayats in the Telugu districts, but the movement on the whole is
almost exclusively Kanarese in its extent. The Vakkaliga of Mysore
correspond to the Kanbi of Gujarat in being subdivided under a general
name meaning simply cultivators. Each of the subdivisions is really a
separate endogamous caste. The principal ones are the Gangadikara, the
Nonaba and the Sada, the second of which is mostly Lingayat, and the
third, Jain. There are other sections either functional, like Halu, the
cowherds, or geographical, denoting immigration. Most of them have
totemistic exogamous subsections. The Pancama and Caturtha Jains and
the Lingayats mostly employ their own priests, but the rest are orthodo.x
in their relations with the Brahman. On the coast of Kanara the land
is held to a great extent by HavTka or Haiga Brahmans, who cultivate
the bitel-palm largely through predial low castes. There are also many

54 5- Ethnography.

cultivators belonging to the fishing and toddy-drawing classes. The chief
caste that can be termed specially agricultural, is the Banta, or warrior,
formerly the rank and file of the militia of the Tulu Chiefs. They have
a Jain subdivision which keeps to itself. The rest observe some of the
Nayar or Malabar customs as to inheritance, and have marriage rules of
their t)wn, which have the effect, it is said, of making the tie "as loose
as it can be". Their neighbours, the Gauda, are probably settlers from
above the Ghiits, where that term is hf>n()rifically used of the headmen of
a village. Further east, in south Orissa, the caste bearing the same name
derives it, apparently correctly, from the Sanskrjt for cow, as they are
of a pastoral character, with traditions of immigration from the north.

The jjrincipal agricultural castes of Telingana are the Kapu. the
Kamma and the Telaga, all of which much resemble each other and come
|)robably from the same stock. The Kapu or Rec.ldi, are widely spread,
though less so than formerly. They are reputed to have more than 800
subdivions, which eat together but do not intermarry. Each subdivision
is in turn split into endogamous sections. Some of the caste own large
estates, earned by military service under the Muslim conquerors of the
14th century, and all are connected in some way or other with the land.
The Kamma, like the Kapu, arc often found in colonies in the south far
beyond the Tclugu country. The Telaga were once a military caste,
and were till recently recruited for the native regiments of the British
army, but now they are cultivators of a moderately high position, and only
differ from their neighbours in being somewhat more fully Brahmanised.
The actual numbers are less than the figure returned owing to the
use of their title by other and probably lower castes out of their native
district. TheKalingi are both cultivators and temple-ministrants on the
Telugu seaboard, with the tradition that they were imported from the
north for the latter purpose before Brahmans had reached Andhra territory.
They wear, consequently, the sacred thread, but are not recognised by
Brahmans as of that order. The rest of the Kalirigi employ their own
priests. They are divided, like the Nayar, into two geographical sections
with quite different customs. A third has had to be formed for the re-
ception of the people expelled from the two others. Their practice is
Brahmanic but their exogamous divisions totcmistic. The Tottiyan are
the descendants of a military body like the Telaga. They were introduced
into the Tamil country, where they are now settled, by the Vijayanagara
Chiefs. As their second title is Kambalattan, probably referring to woollen
blankets, and their subdivisional titles being also those of a pastoral
character, it may be inferred that their original occupation was that of
shepherds. Locally they are much dreaded for their magical powers, but
in compensation, their cures and charms for snake-bite bear a high re-
l)utation. The name VellaUn, in the Tamil country, corresponds in its
generality with that Kunbi or Casa in other parts of India, and merely
implies a cultivator. The wide diffusion of the community so called prevents
it from being a caste, in the sense of a homogeneous body, as irrespective
of the four great geograi)hical sections, over 900 subdivisions were re-
corded at the census. By careful filtration, the number was substantially
diminished; nevertheless, the residue is very large, and owing to the
accretions from lower castes as they rise in the world, it is constantly in-
creasing. It is unnecessary to point out that in such circumstances the
endogamous sections are many and minute. Of the main divisions, that

Castes and Caste-Groups. B, The Village community. 55

called the Tondamandalam, of the old Pallava kingdom, round Arcot,
stands highest. It settled in its present location in the 8th century A. D.,
and is strictly Rrahmanistic in customs and religion. The Kongu, on the
other hand, who are found in and about Coimbatore, arc so far below
the rest that none of the other subdivisions will eat with them, and they
are sometimes considered a separate caste, under the name of KavaiK.lan.
Apparently, ti)0, their marriage regulations have not passed away from
the old Dravidian type, and Brahmans are not employed, as they are
amongst all the other Vellalan bodies. In the jMalayalam tract, below the
Nayar, ]\Iappila and Nambutiri Brahman, the cultivating castes belong to
bodies having other traditional callings, or are field labourers who have
occasionally got hold of a small estate. They will be found, therefore,
under their respective headings in later paragraphs.

ij 37. Specialised cultivators (5,968,700). The majority of the castes
coming into this category are branches of the great agricultural bodies,
separated from them, as stated above, in view of the inferiority in rural
esteem of the produce they cultivate as compared with cereals and
other crops grown on a large scale. Thus, the AraTn are of the same
stock as the Kambo ; the Mali, Kachf and Murao, are all derived from
the Kurml, and the Saini belongs to the Mali. In contradistinction to the
growth of roots and vegetables, the care of the bitel-vine has no disgrace
attached to it. This may be partly due to the use of vegetable manure
only, and partly, no doubt, to the consideration that the presentation of
a little packet of the leaf with areca nut is an important formality in
social intercourse. In the greater part of India the bitel-vine is grown by
a special caste called Baraf, Barui or Bari. The last title, however, is
only used south of the Vindhya, and in the north is applied to a lower
caste of different occupation. Apart from linguistic distinctions, the Barai
is much subdivided into endogamous sections, and most of them hold a
good position in society. In the Dekkan and Karnatic there is a small
caste of Brahmans, the Tirgiij, who have taken to growing the bitel-vine, and
the Bari are said to be immigrant from Central India. In the Tamil country,
the Scnai kkiidaiyan do what most of the Barai avoid, that is, sell the
leaves themselves, instead of making them over to another caste for the
market. This caste has the further peculiarity of belonging to the Left-
hand in the local distribution, thereby grouping itself with the artisans,
a position which does not, however, militate against its respectability, or
prevent the Brahman from sharing with Vellalan the priestly ministrations
required in the caste. The Kodikkal, another bitel-vine growing caste
is only a subdivision of the VeOajan, based, apparently, upon its occupation.
As the areca-palm only flourishes in certain localities, its cultivation is
undertaken by the ordinary agricultural classes. Reverting to the market
gardener, the Arain of the Panjab is a true caste in the north and
east of the Province, but in the west the title is purely occupational,
like Jat in the same tract. The community seems to have come up the
Indus from Multan or north-west Rajputana, and settled along the Ghaggar
river, then probably of an irrigational capacity it has long since lost.
Thence they spread across the Jamna into Rohilkhand, and northwards
into Jalandhar, which is still one of their principal seats. Here they are
not only garderncrs but general cultivators of considerable reputation for
skill and industry. They are, as stated above, akin to the higher caste
of the Kambo, but with a far greater inclination to accept Islam. The

56 5- Ethnography.

Miliar of the north-west, who arc entirely Muslim, are lower in position
than the Arain, though they appear from the names of their subdivisions
to be a branch of that caste. The Mali get their name from the garlands
it was their mission to prepare for the decoration of the temple deities
and ti) throw round the necks of honoured guests at social ceremonies.
They have long branched out into all kinds of garden cultivation, and
their numerous subdivisions are frequently based upon the produce to
which they are respectively devoted. Those who grow flowers, for instance,
do not intermarry with the vegetable -growers, and the latter draw a
distinction between themselves and the branch wliich grows onions, turnips
or turmeric. The KachI has taken in upper India to the poppy and le-
guminous edibles, leaving roots to his poorer relative the Murao, who
is said to take his name from the radishes he grows. Some sections of
the Kachf, again, abstain from cultivating the sugarcane or chillies. The
Sainj, another branch of the Mali, are found in the east Panjab and in
Rohilkhand, where they are as much general cultivators as gardeners. In
the former tract a good many of them are Sikhs, but the more prosperous
claim Rajput blood. They stand high in their calling and seem to be living
down the taint of the garden. In the Peninsula, south of the sphere of the
Mali, the only specialised cultivator in addition to those already mentioned,
is the Ti gal a, now located in Mysore and the south Dekkan. This seems
to be one of the few castes which have moved northwards from the Tamil
country, but they have retained neither the customs nor language of
their origin.

t; 38. Cattle-breeders (11,965,500). These are taken next to the agri-
culturists because they occupy a very similar social position, and also
because, with the expansion of tillage, the grazing area is getting restricted
and a good many of the formerly roving castes have settled down to
cultivation. The prominent place assigned to cattle in the Suktas and the
universal veneration of the Brahmanic community for the cow bear testi-
mony to the antiquity as to the honourable character of the calling, and
in upper India the cattle-breeder ranks almost as high as the cultivator.
This is not invariably the case, however. The wandering life arouses
suspicions of unorthodox feeding and intercourse generally. Then, too,
the use of the ox in agriculture now vies in importance with that of the
cow in domestic life ; but the supply of the indispensable bullock cannot
be kept up without surgical operations repugnant to the conventional
notions of jnirity and respect for animal life. Furthermore, the supply of
milk for the home is, by all Vedic tradition, commendable, but the sale
of dairy produce as a trade entails relegation to a lower position. In old
times, however, the Abhira, or cowherding tribes, were powerful in the
Satpura, the south Ganges valley and even the lower portions of Nepal,
and founded dynasties which were overthrown by the Gond in the first-
named traC; and by the Kirata in the last. The leading tribes seem to
have been of western origin, and are supposed to have entered India long
after the Vedic Arya. In upper India they go by the name of Ahir, derived
from the Abhira just mentioned, a term which was applied by some
Sanskrit authors to all tribes of the lower classes throughout the north-
west. Under this name they are spread in considerable numbers all over
Rajputana, Malva, the south-eastern Panjab, the upper Gangetic valley
and Bihar. To the east, the lack of wide stretches of open pasture has
prevented the formation and maintenance of a strong and well-organised

Castes and Caste-Groups. B. The Village community. 57

pastoral community, so a number of distinct and generally not very large
subdivisions are grouped under the general title of Goala, recruited from
many local castes of lower origin than the pastoral bodies of the north.
Many of them, too, are as much agriculturists as cattle-breeders. The
same may be said, also, of the Ahir themselves, in the Panjab, where
they are amongst the most successful and enterprising cultivators of the
Province. They have never, it is true, achieved a dominant position any-
where in modern times, but the Jat and Gujar treat them as equals, ex-
cept, of course, in regard to intermarriage. According to the Census,
about half the total number of Ahir are found in Agra, Oudh and Bihar.
They are said to have migrated to these regions from the plains of Kach,
west Rajputana and Kathidvad. Assuming their connection with these
parts, especially the last named, a basis will be found for their invariable
assertion in the Gangetic region that the cradle of the Ahir is Mathura.
Few legends are more wide spread in India than that of the dalliance
of the most popular of Puranic deities, Krsna, with the Gopi, or milk-
maids, of the Vraj district: and the Jaduvansi line, headed by Kr.sna himself,
found its second home, after its expulsion from Mathura, at Dvarka and
in the north of Marvad, the very tracts inhabited by the Ahir before they
entered Hindustan. Traditional descent from the IMathura Jaduvansi is not,
however, confined to the Ahir of the north, but is claimed by the Gaura
and other cowherds of Orissa, and even by some far to the south of the
Arja pale. Except in the Panjab, the Ahir enjoys but a poor reputation
as a husbandman, though everj'where he is admitted to be company for
the higher peasantry. This, however, may be, as in the case of the Gujar
in those parts, a question of policy, with a side-glance towards the village
cattle, which are too apt to stray into the Ahir's herd without their rightful
owners' knowledge or consent. The Gaul i of the west Central Provinces
and north Dekkan, is the descendant of the tribes which, as just mentioned,
once ruled the Satpura from Khandesh and the Sahyadri, to near Saugor,
and were only expelled by the Gond in the i6th century. As they are
mentioned in the Nasik cave inscriptions, they must have been long
established in their dominion. Alongside of them is the Govari caste,
which has no trace of immigration either in nomenclature or tradition.
In the Chattisgarh country, to the east, comes the Ravat, another cattle-
breeding caste of long standing in that region. The two last mentioned
castes which in 1 891 numbered about 350,000 persons, do not appear at
all in the returns for 1901, so they have probably been compiled under
Ahir or some other general title. Two other cattle-breeding castes of
upper India may be mentioned, the GhosT, an offshoot of the Ahir, or
as some think, of the Gujar, which has been converted to Islam. They
occupy a comparatively low position, and near the large towns confine
their attention to the dairy side of their occupation. The other caste is
the Rabari of Rajputana and the Gujarat peninsulas. They are of Marvad
origin, but wandered to the coast, and now breed both cattle and camels,
and some of them even become shepherds. In the north they confine their
trade to camels. In the Dekkan, the Gauli, and further south, the Go 11 a,
represent this industry. In the Tamil country, the cultivator generally
breeds his own cattle, and only one caste devoted to this occupation
appears in the return. This is the Kannadiyan, a small body, of apparently
upland origin. The Golla of the Telugu and Kanarese tracts, are thoroughly
local castes, but, having become Brahmanised, cast back to Mathura and

58 5- Ethnography.

the Gupi. Most of them are settled in villages, but f)nc section, in Mysore,
is still nomadic during the open season, and does not intermarry with the
others. In Mysore it used to be the duty or privilege of the Golla to guard
State treasure in transit, and the official now responsible for sending off
the remittances is still occasionally called by that name, albeit he may
be a Brahman or Muslim.

Sj 39. Village artisans and servants. Handicrafts and mechanical
arts have always held a low place in jiublic esteem in India, and to this
day, in societies moulded on archaic lines such as those of the lower
Himalaya, the division between them and agricultural occupations is very
marked. An exception is found, as a rule, in the worker in the precious
metals, a trade tolerated, if ncjt honoured, even in Vedic times. Throughout
the greater part of India the castes of the artisans are graduated according
to the material used in the calling.

a I Combined crafts (1,263,900). From at least the date of the Ma-
habharata, five trades, called the Pafickalsi", stand out from the rest,
and arc usually grouped together. The goldsmith comes first, except in
Bengal. Then comes the brass and coppersmith and next the carpenter
or other worker in wood. The blacksmith follows in a lower place, partly,
no doubt, because his is a dirty calling, partly because he has to use
bellows made of oxhide, and partly, again, because the metal in which
he works is black, the unlucky colour. In the Gangetic valley, too, there
may be some association between the village and the nomad blacksmith,
who is probably of K61 origin and shares the reputation of the gipsy tinker
and farrier of Europe. The fifth place in this hierarchy belongs to the
stone-worker, which, exept in the south, is a more modern and probably
a purely functional body. The above castes are not always strictly separated
in occupation: sometimes the carpenter becomes a blacksmith, and the
masonry, like bricklaying is done by an outsider; the latter being held to
verge upon the task of the potter, which is impure. In the Dravidian country
the five are found merged in a single group, called the Kammajan in Tamil,
Karhsala in Telugu, andPancala in the Karnatic. The occupations then
fall into subdivisions. This cohesion seems to have been promoted, if not
initiated, by sectarian influence. It appears that in this part of India the
artisans used formerly to be excluded from the main village site, and forced,
like the leather-workers and scavengers, to live in hamlets of their own,
detached from the rest of the community. As their work grew in impor-
tance, their origin, which was probably amongst the servile classes, tended
to be forgotten or ignored, and they were admitted within the walls, and
allowed certain privileges in the way of social display which had before
been reserved for the higher classes. Then followed the great Southern
schism of the Right and the Left-handed castes, in which the artisans
arrayed themselves en masse against the Brahmans and few others. It is
now gencr.'illy held that this movement arose out of the levelling doctrines
of the Buddhists or Jains of the south, which had been largely adopted
by the lower classes; but whether the artisans, thus encouraged, led a revolt
against Brahmanical authority, or whether, on the decline of Buddhism,
the Brahmans took this means of setting the schismatics back into their
place, is not certain. In the present day, the differences between the two
factions, which are acrimonious and often turbulent, arise, not out of doc-
trinal questions, but on points of what may be termed processional privi-
leges, such as the right to have the marriage-escort preceded by drums

Castes and Caste-Groups. B. The Village community. 59

and trumpets, to have a mounted convoy in attendance, to carry certain

emblems of a quasi-religious signification ; above all, to exceed a con-
ventional maximum number of pillars to the marriage-booth. Castes whose
technical skill and circumstances have raised them far above the class
from which they sprang have often shown the tendency, as stated in
an earlier section, to embrace a new scheme of reform which combines
religious doctrine with the weakening of the barriers which prevent their
equivalent rise in social position, and in this case the democratic teachings
of Jainism and Buddhism had the further backing of the propaganda of
Basava in the north Karnatic, with the result that most of the Paiicala
became Liiigayat and, therefore, anti-Brahmanist. None of the Five grouped-
sections employs Brahmans or acknowledges the authority of that order,
and all ceremonies are performed by priests of their own body. For some
time past the Panckalsi have claimed descent from Visvakarman, the He-
phaestos of the Brahmanic pantheon, and call themselves Visva Brahmans,
assuming all the attributes of the sacerdotal order. In this respect the
Southerners do not stand alone, since a similar claim is put forward by
various artisan castes in other parts of India, especially by the goldsmiths.
It is needless to say that whatever title or practice may obtain currency
within the community, its sanction by the outside world has to be secured
through the Brahman, who naturally will have none of it. Authorities differ
as to the homogeneity of the Panckalsi. By some it is said that the occu-
pations are interchangeable, and that families or individuals pass from one
to another without any alteration of social status or loss of right of inter-
marriage. Others say that in the Tamil country- the divisions do not ge-
nerally intermarry, but that this is not the case in the Telugu country-,
where all five certainly eat together, and are said to intermarry. The
Kanarese branches follow the rules of the Liiigayat community. In the
Malabar tract the five stand on a different footing, and take a far lower
position. They are amongst the impure castes and do not employ their
own people as Brahmans. The As'ari, or carpenter, who is the house-
builder of the coast, stands above the rest, and at the ceremonies con-
nected with the erection of a building he is allowed to wear the sacred
thread. The Tattan (goldsmiths 1, Kollan (blacksmiths), and Mus'ari
(coppersmiths), intermarry. The stonemason is not an important coast artisan,
but above the Sahyadri and in the south, the number of stone temples
and images is so large and their use so ancient, that the functions of the
stone-worker have always been in great request ; so much so, that in
some of the inscriptions this craftsman is invested with the title of Acarya,
or teacher, which though the Panckalsi nowadays use it of each other, is
not ordinarily conferred on any but religious or literary instructors. In
consequence of the use of the general title Kammalan instead of the sub-
division, it is impossible to give the numbers of the Panckalsi exercising
the respective trades included under it, except for the comparatively limited
population of the Malabar coast, and this, irrespective of the peculiar
constitution of the community, is a reason for dealing with the latter
apart from the corresponding castes of the rest of India. There is, how-
ever, in Bengal, a somewhat similar grouping in the case of the Kamar
or metal-working castes. This body apparently started with a variety of
functional groups of different origins, and is now welded into a sort of
caste, subdivided according to the metal used, and bearing the general
title usuallv given elsewhere to the worker in iron. The legend in which

6o 5. Ethnography.

the Kamar trace their descent from Visvakarman, indeed, is very much
the same as that by which the iron-smelting Asura of the Kol race justify
their origin from the same ancestor, thus confirming the general view
as to the non-Aryan foundation of the caste. The social graduation of the
subdivisions is curious, in that the worker in iron stands first, and inter-
marries only with the worker in brass, and the bell-metal craftsman stands
above the goldsmith. The latter, indeed, under the name of Sckara, or
Svarnakiir, though he holds himself higher than the wealthy Subarnabanik,
mentioned along with the Traders, must have something against him from
days of old, as the Brahmans which serve his subdivision are not in com-
munion with the rest of their order, whilst those who perform similar
functions for the rest of the Kamar are under no such interdiction. The
Niyariya, or Dhuldhoya, is a parasitic caste upon the Sonar, and lives
by extracting the gold out of the refuse of the latter's shop. He is usually
allowd to be Sonar in blood as in occupation, but in the north is often a
Muslim, even when the goldsmith is Brahmanist.

b) Gold and silver workers (1,290,500). The goldsmith is very often
a pawnbroker and money-lender as well as a manufacturer of the orna-
ments which constitute the main capital of the peasantry and indeed of
most Indian middle classes, and in both capacities has acquired a very
indifferent reputation for straight-dealing. According to one popular saying,
he so regretted having made a nose-ring for his own mother without
sufTiciently adulterating the metal that he cut her nose off to recover it.
In the Gangetic region the caste, which is subdivided to an astounding
extent, is said to be a composite one, but still holds a position superior
to that of the other artisans. It is said to be clo.sing up its ranks, too,
and forming large endogamous sub-castes out of its numerous minute
exogamous sections. In this tract the Sonar does not seem to be putting
forward the same pretensions to be Brahman that he does further south.

c) Carpenters (2,688,100) and d) Blacksmiths (2,362,300). It is the
Lohar and Rarhaf, who refer themselves back to Visvakarman, and who
have a joint sub-caste called Ojha claiming to be Brahmans, not apparently
without a certain degree of recognition, though not to the full extent of
their desire. In the west, the Sutar, or carpenter, throws back to the Gujar
or Vania, and in the Dekkan, to the inevitable Visvakarman. The Lohar
seems everywhere constant to the latter. There seems to be a general
tendency to make these two functions interchangeable even though the
castes remain distinct. In the Maratha districts, both above and below
the Sahyadri, the Sutar does the village ironwork, consisting mainly of
simple repairs such as retyring cart-wheels or reshoeing the plough and
so on. In the western Panjab it is the same. In the east of that Province,
the Tar khan and the Lohar arc the same caste by origin, but the car-
penter stands higher, and when both occupations are followed, sub-sections
are formed which do not eat together or intermarry. There is also a body
of Lohar in the south, along the Rajputana border, consisting of Rajputs
who, from stress of circumstances, probably famine, were driven to adopt
this means of getting their living, and though called Lohar, are ai)art from
and above the rest. The Khati, again, is both carpenter and blacksmith
in some parts of the north, ranking with the former, but along the Jamna
the caste is wheelwright, and considered a subdivision of the Barhai.

e) Masons (51,400). The Thavi of the sub-Himalayan region, is an
offshoot of the carpenter, but, as the dwellings in those parts are chiefly

Castes and Caste-Groups. B. The Village community. 6i

of stone, the caste has developed into masons as well as workers in wood,
and in the plains, too, the Raj, when the title is not merely functional, is a
carpenter turned mason. The large caste of the SutradhSr in Bengal, is of
local origin, probably akin to the Kaibartta, but is now much subdivided
into functional groups taking rank a good deal according to the nature
of their work, such as boat-building (one of the lowest), wheelwright,
builder, turner and painter, all independent of each other. Some have
established a body of priests of their own. The barber, whose function
is one of the touchstones of rank, considers them high enough to be
shaved by him, but will not undertake their pedicure. This discrimination
between the different branches of the craft is found elsewhere. The car-
penter who undertakes the repair of municipal conservancy carts, for
example, has, for an obvious reason, to sacrifice his position ; and the
making of oil-presses and, as just mentioned, boat-building, is considered
degrading, owing, probably, to the indirect connection of these articles
with the destruction of life. Both carpenter and blacksmith belong to the
class of village artisans remunerated by customary shares in the year's
harvest. During the cultivating season, therefore, they are bound to de-
vote their time to the needs of their clients, but during the rest of the
year they make carts, bedsteads, irrigation-wheels, and other articles which
are charged for in the ordinary way, at a price either cash or kind, more
usually the latter. The workers in brass and copper appear among the
Panckalsi, and can claim considerable antiquity, but they are urban rather
than village castes, and are rarely found, except casually, in any place
smaller than the local market town. At the same time, their occupation
enters largely into village life, since there is no more distinctive mark
of the prosperity of a tract than the substitution of metal vessels, especially
of the larger sorts, for the porous earthenware which was formerly in
universal use. Once established, the demand for the former is extensive,
as each family requires its own complete set, to obviate the risk of con-
tamination by contact with other castes.

f) Brass and copper workers (206,800). The manufacture and provision
of these articles are in the hands of the Kasera and Thathcra castes
in upper India, and in those of the cognate bodies called Kasar, Kansara
and Tambat, in the west, and Bo gar <:ir Kannan in the south. In the
Karnatic the Caturtha and Pancama Jains have a good deal of this trade
in their hands. In the north they hold a better position than in the south,
having traditions of Banya origin. In the sub-Himalayan tract, however.,
they belong to the earlier and darker tribes. They seem to be, on the whole,
more homogeneous than most castes, possibly because their trade has
fewer ramifications, and they do not deal, as a rule, in the articles they
make, but dispose of them to special traders for sale to the public. At
the periodical gatherings at the great centres of pilgrimage, the booths
of the brass and copper vendors are well to the fore in the fair which
is always held as a subsidiary attraction on such occasions, and as the
wares are conveniently portable, the business is brisk. The mason, which
is the last craftsman to be dealt with under this group, does not, in most
parts of upper India constitute a real caste, but belongs to a functional
group recruited either from the carpenter and lower menial castes, or
occasionally from others, whose members have been driven to manual
labour, and selected the branch which is least associated with impure ma-
terials. There are, however, true castes of this trade, such as the Gaundi

62 5- Ethnography.

and Ka(,li6 of the Dekkan and Gujarat, who have lived down their pro-
bably pre-Aryan descent. The stoneworkers of the south and some of the
masons, largely consist of members of the salt-working castes whose oc-
cupation, since the manufacture of salt was undertaken by the State, has
been seriously restricted. In Gujarat, the caste has been formed by se-
paration from the agricultural labourer, and in parts of the Gangetic valley,
from the lime-burners and manufacturers of saltpetre. The making of bricks,
owing to the impurity t)f the material used for the kiln, rests with the
Kumbhar, or Potter caste, which comes into a later group.

§40. Weavers (9,541,000). The people of India were wearing cotton
garments in the days of Megasthenes, and do so still. No wonder, there-
fore, that the occupation of hand-loom weaving is one of the most widely
distributed in the country, and forms the traditional calling of castes
containing nearly ten millions of people. In its palmy days the craft
reached a wonderful pitch of skill and refinement, especially under the
patronage of the Delhi Court, which monopolised the whole of the Dacca
output of "flowing-water", "gossamer" and other choice muslins, the art
of weaving which has long been lost. Even the staple everyday fabrics
made far beyond the imperial ken, at the seaports of the gulf of Cambay,
the Malabar and the Coromandel coasts, always found a ready market in
Europe and the Levant. The weaving community seems, nevertheless, to
have been anything but prosperous. Before the end of the i8th century
they were reported by British officials to be "a timid and helpless" folk,
and even then, were, as recent experience has proved them to be still,
among the first to feel the pinch of famine, when a wide-spread failure
of crops reduced or stopped the purchasing power of the peasantry. Since
then their market has been seriously curtailed by the competition of
European machine-made goods, and it is only in the coarser lines of
material that they hold their own. The weaver is not one of the menials
who is, so to speak, on the village staff: that is, he is not entitled to a
customary share of the harvest, but is paid for what he makes and sells.
With one or two exceptions, the weaver castes occupy a low position,
considering the character and utility of their function. This is doubtless
due to the fact that the latter originated amongst the pre-.Aryan races,
who subsequently became the helots of those to whom cotton was unknown
before they exchanged the steppes of the north for the more genial tem-
perature of sub-tropical India. The weaver, though below the peasantry,
is far above the village menials who do field-labour and work in leather
and other impure materials. He represents, in fact, the highest rank to
which castes of that origin can attain. Perhaps the best instance of this
position is found in the Tanti of Lower Bengal, who enjoy a rank much
above that of any other weaving-caste, and even, intermarry, when suf-
ficiently wealthy, with castes like the Kayasths. In their case, however,
there is no question of evolution from any lower Deltaic tribe. It is not
known whence they came, but the country in which they are now found
is not a cotton-growing tract, and the weaving industry, accordingly, was
probably introduced from the north-west, the origin of the craftsmen being
obscured by promiscuous recruitment, and condoned in consideration of
their skill and utility. There are other cases of weaver castes of superior
position, such as the Khatrl or Patve of Gujarat and Central India, w^ho,
from the beginning dealt with no fabric but silk, and the probably kindred
caste of Pattunurkaran, in the Tamil country, which found its way by devious

Castes and Caste-groups. B. The Village coMjruNiTY. 63

routes and with many halts, from Malva to the south. But the mere re-
striction of their operations to the more valuable products is nut, of itself,
enough to raise the caste above its fellows in the eyes of the world, for
the Tantva of Bihar, who are silk-workers, but also breed the worm, rank
far below the Tanti, who use cotton. On the other hand, the handling of
jute or hemp seems of itself to keep a caste to the bottom of the craft,
as in the case of the Perike and Janappan of the Dravidian country,
the Kapali of Bengal, and the Dhor of the Dckkan. In regard to the
evolution of the weaver from the servile castes, a good instance is found
in the east of the Central Provinces and the adjoining Orissa hills, where
the process is still going on. The Panka, a tribe of Kol or Dravidian
origin, with its cxogamous totemistic structure, does the coarse weaving
of the tract, and also cultivates, either as an occu])ant or a field labourer;
but in many villages it is not admitted within the site, and has to dwell, like
other impure menials, in a detached hamlet. In the Central Provinces the
Panka has joined the Kabirpanthi sect in considerable numbers, like the
leatherworking castes of the neighbourhood, with the further inducement
that the founder of the sect was himself a weaver. The Ganda, another
weaving caste of the same region, but mostly inhabiting the ])lains, is
closely related to the Panka, and, indeed, is often held to be a subdivision
of the latter; but its members are now not weavers so much as cultivators,
village watchmen and drummers, nor do they share the Kabirpanthi views
of the Others. To the south of these castes, acros^ the hills, are the Domba,
a tribe of hill weavers, low in their habits and trade-skill. They mostly
belong to the ISIadras territory, but, from their name, it is possible that
they may appertain to the great Dom tribe of the north of the Ganges,
members of which are found detached in the Dekkan and Karnatic. Like
the Panka, they are classed with the lower menials of the village, and
perform the same unhonoured functions. In nearly all the other parts of
India the differentiation of the artisan from the menial has been more
definitely carried out. The Kori, the chief Brahmanic weaving caste of
Upper India, together with the Julaha, the corresponding division of the
IMuslim, are now (juite detached from the leather-working caste from which,
according to the nomenclature of their subdivisions, they sprang. In the
case of the Julaha, the sectional affix is falling into disuse, and with it
the customs with which it is associated. The Kori adhere more closely
to their ancestral practices, possibly because the chances of rising in
position in the Brahmanic world are not to be compared with those
offered by Islam, as embodied in the popular saying — "Last year I was
a Julaha (or Nadaf); this year, a Saikh, and next year, if the harvest be
good, I shall be a Saiad". Both castes work chiefly in the coarser fabrics,
as they have been hard hit by foreign competition in the finer class of
weaving. Some of the KOri sections are of the Kabirpanthi sect, but others
pay their respects to both the orthodox Brahmanic deities and to the
popular Muslim saints of the locality, a practice reciprocated by the Ju-
laha, who worship Mata Bhavanf, where she holds the popular favour.
The Julaha of the cities have the reputation of being a specially factious
and quarrelsome body — "Eight Julaha fighting over nine hukkahs" —
say their neighbours. The place of the K(>ri is taken by the Balahi in
Rajputana and Central India, a caste allied, like the rest, to the Camar,
or leather-worker. In southern India the weaver castes, though varying
in rank, seem to have long acquired a higher position than in the north.

64 5- Ethnography.

The Kaikkulan, or Tamil weavers, share, it is true, an ancestor with the
Paraiyan or menial caste, and used to be relegated with the rest of the
Kammala with whom they were classed, to a detached hamlet. By dint
of clean living, however, and the employment of Brahmans, they now
occupy a respectable position. Most of the other weavers of this part of
India are of Kanarese origin. A good many are returned simply under the
general title of Neyige, the Mysorean term for weaver, and are probably,
like the Sale of various subdivisions, very largely Liiigayats. The Sale
have long been settlers to some extent in the Tamil country where they
wove silk with much profit, but lost ground under the competition of the
still more skilful Pattunurkaran. In the Dekkan and Central Provinces they
are found in different grades, according to whether they work only in white
or add a border or fringe of coloured silk. The Devaiiga and the Togata
are other sections of the Kanarese weaving community, lower in position
than the above. The Togata, indeed, are not found in their native country
at all, but have permanently settled in the south. A caste of Bengal
weavers, the Jilgl, has been mentioned in connection with the ascetic
body of a similar name. Its origin is unascertained, but it is not affiliated
to the leather-workers. Its low position may be partly attributed to the
pretensions it has made to higher rank, thereby entailing an unusual con-
centration of Brahmanic displeasure. Though suffering like its fellows from
European competition, the caste till recently had stuck fairly closely to
its traditional calling. The Ko.sti of the Maratha country h'olds, like the
Kaikkolan, a middle jjlace between the silk-weaver and those of servile
origin. Brahmans are em[)loyed in the caste ceremonies and the Ko.sti
lives, as a rule, very like the poorer Kunbi. The famines of recent years
caused much distress amongst this caste, and, from their sedentary- life,
it was difficult to adopt means for giving them fitting relief work. They
are endeavouring to evade the results of foreign competition by weaving
British yarn, whereby they produce a fabric which combines fineness with
the strength and durability of hand-loom work.

§ 41. Oil-pressers (4,517,600). Wherever oil-yielding seed or nut is
grown there is an oil-press in every village of average size. The material
most extensively used in the interior is sesame, with linseed and the
castor-bean for burning. Along the coast the coco-nut is the chief oil-
producing material. The castes engaged in oil-pressing do not everywhere
take the same social position. Generally, their rank is low, because the
occupation is undeniably a dirty one ; but there arc degrees even in im-
purity. In most parts those who only press sesame, or oil used in cookery,
are higher than those who prepare the oils used for burning or lubri-
cation. But sometimes a distinction is drawn between those who get out
the oil by boiling the seed and the majority, who use the press. Amongst
the latter, in turn, those who yoke two bullocks to the press take prece-
dence ovei those who use only one, and the subdivisions are named ac-
cordingly. In the present day, however, the single bullock is the rule, and
this blindfolded and unfortunate agent is everywhere the proverbial type
of dull and endless toil. Finally, the oil may be allowed to drip through
a hole in the press or may be baled out of the receiver with a little rag-
mop. In parts of Bengal the latter process alone is honourable, the reason
being that when oil i)rocured by the former was presented to the goddess
Bhagvatf, she drew a trenchant and celestially outspoken analogy between
the form of press and the human body, in token of her disapproval of

Castes and Caste-Groups. B. The Village community. 65

the method adopted. Hence, the Tcli who mops out his oil will have
no intercourse with the Kalu, though both are subdivisions of the same
caste. In the Panjab the Tcli is Muslim, and one of the divisions has
separated into a distinct body, the Oasab or butcher, both ranking with
the Julaha. In other parts of upper India, the Brahmanist Tcli is resiicc-
tablc, but on a low plane, and some, including those of Bihar, are served
only by Brahmans who are out of communion with their fellows. In
Bengal, Gujarat and the Dekkan, the oil-prcsser is often a grain-dealer
or shop-keeper, and in the first named province attains to considerable
wealth and importance. In the Dravidian country the caste is known by
the name of the oil-press, Sekkan or Vaniyan, in the Tamil districts,
and Gandla, Ganiga, or Jotipan, in Tclugu and Kanarcse. The Telugu
and Tamil castes employ Brahmans, wear the thread and generally follow
the customs of the upper castes of cultivators. The Kanarese castes are
more subdivided, but employ Havika Brahmans when available. Some are
Lingayats. The oil-presser in Malabar stands on a different footing to the
rest. In the northern region he is ranked with the impure, and kept
down. In the south of the tract, however, he is one of the castes
which has crejjt under the comprehensive title of Nayar. In neither case
do the oil-pressing castes wear the sacred thread as they do above
the Sahyadri, nor do they employ Brahmans. The trade is one which
has suffered considerably of late from the competition of mineral oil for
burning purposes, and numbers of the Teli are taking to cultivation for
a living.

§ 42. Potters (3,521,8001. The Potter is one of the recognised village
staff, and in return for his customary share in the harvest is bound to
furnish the earthenware vessels required for domestic use. His occupation
goes back to the time of the Vedic Suktas, and varies in its demands
upon the worker according to the customs of the province or tract, the
consumption of earthen platters being in some parts enormous, whilst
elsewhere metal is substituted, except for water and storage. The po-
sition of the Kumhar, Kumbhar, or Kus'avan, is above that of the helots,
but is undoubtedly low. This is made manifest by the association of the
caste with the donkey, the saddle-animal of S'itala, the goddess of small-
pox. The Dhobi, or washerman, is the only other of the settled or
village castes which makes use of that useful, but in India foulfeeding,
animal. Where the caste is much subdivided those who use the bullock
for carriage are superior to the patron of the humbler animal. Those who
work on the wheel, again, do not intermarry with those who use a mould
or make images. Elsewhere there is a distinction drawn between the
artificer who only makes large vessels, and accordingly stands to his work,
and him who squats on the ground. As in the case of the weavers and
oil-pressers, the Bengal potter seems to enjoy a better position than his
comrade of upper India. In Madras, too, both Telugu and Tamil Kusavan
wear the sacred thread, and some sub-divisions employ Brahmans, as in
Bengal, whilst others have priests of their own community. Where bricks
are in use the potter undertakes the kiln, and though, as above stated,
he has to use fuel collected from sweepings and other refuse, he is not
called upon to touch the lowest kinds of filth, and escapes therefore the
condemnation inflicted upon the scavenger. His donkey, too, where it is
in general use, is employed when the kiln is not in operation in carrying
grain and other produce. In most parts of the country, the potters some-

Indo-Aryan Research. II. b. 5

66 5- Ethnography.

times hold land, and in others take service in large households. In the
Telugu country they are even in request as cooks, one of their traditional
occupations in that region.

5; 43. Barbers (3,698,300). Shaving and the paring of nails are
important parts of many Brahmanic ceremonies. The arrangement of mar-
riages is the work of an expert and trustworthy go-between; the formal
communication of domestic occurrences (except deaths), the provision
of music before processions, the accompanying, with a torch if necessary,
of distinguished strangers on their arrival in the village, together with
the essential function of gossip, all these qualifications and duties go to
make the barber a much esteemed member of the village hierarchy, on
a regular annual stipend either from the individual householder or out
of the land or its produce. The Nai, Napit, Ambattan, Mangala, or
Hajam, moreover, is usually the only person in an average village with
any knowledge of surgery, though other castes can come to the rescue
of a person afflicted by such ailments as are known to yield to charms
or spells. It is this practice of surgery, it is to be feared, which relegates
the Barber to a social position much below the esteem he enjoys as an
individual. The caste, however, as a whole, is exclusive and particular.
In some tracts of the west, each caste has its own barber who will attend
to no other. Everywhere, too, there is a social limit below which a barber
will not shave. Nor, though his mediation is essential to the announcement
of good tidings in a formal manner, will he ever consent to carry round
the news of a death, a duty which is imposed upon a caste which is
presumed to be below the bad luck likely to accrue from so doleful a
task. In most parts of India except the Panjab, where the Jhinvar's wife
takes the office, or where a Camari is employed, the barber's wife is the
midwife or monthly nurse, and occasionally she acts as hair-dresser and
manicurist to women. In Bengal, the latter occupation is alone the custom,
and that but rarely. Indeed, the position of the caste, as well as that of
the Bhani.lari, the barber caste of Orissa, is much better in the east
than in other parts. An exception must be made in favour of the ^larayan
of the IMalabar coast, who in the north of the tract is the barber of the
Nayar, but as the south is approached, sheds his occupation to some
extent, and acts as drummer generally, and as Nayar priest at funerals.
Still further down the coast, the work of shaving is left to a caste called
Velakkattalavan, but which calls itself Nayar. Meanwhile, the Marayan
have passed into temple-service, drumming and the conduct of funerals,
and give themselves the name of Attikuricci or Ambalavasi. Under this
transformation, the caste ranks next to the Brahman, and will not eat
with Nayar: but no more will the Nayar eat with the Ambalavasi. The
Mangala are the barbers of the Telugu districts, but as their connection
with preparing the mourners for a funeral renders that name unlucky,
they arc usually addressed as Bajantri, or musicians, in reference to the
other branch of their profession. The barber is everywhere credited with
vast experience of the outside world, together with a quite exceptional
accjuaintance with the esoteric affairs of all the families in his village.
The Brahman, therefore, ministers to him without reluctance, and what
with fees, presents, feast offerings and other emoluments, he often acquires
quite a well-to-do position and is respected accordingly. There are as
many proverbs about him as about his confrere in the West, and both he
and his razors are mentioned in the Suktas of the Rgvcda.

Castes and Caste-Groups. B. The Village community. 67

!< 44. Washermen (2,887,600). In the south and west of India, the
washerman is generally jjlaced next below the Barber castes, but in Agra,
Oudh, Bihar and Bengal, his position is far lower. This difference arises
from convention and custom. In the one region, all but the wealthy do
their own washing, either in person, at the tank in the mornings, or
through the women of the family. In the north and east, however, the
handling of soiled clothes is a polluting task, and the Dhobi ranks no
higher than the leather-worker. He is moreover associated in these parts
with the donkey, like the Kumhar, and pays the penalty of the convenience.
In most parts of upper India, in Bengal and in the Panjab and parts of
the Karnatic, the washerman is one of the hereditary village staff, and
gets his share of the crops like the artisans. In Bengal he has even to
take a jiart in the marriage-rite of the superior castes, a function which
he is not called upon to perform elsewhere. At the same time, it is
usually a lucky omen if on leaving home one catches sight of a Dhobi
in clean clothes. The last qualification is of uncertain signification. It
may be due to its rarity, or, again, it may be connected with a popular
saying that the Dhobi's outer garments belong to his patrons. Except,
however, in the localities just named, the Dhobi belongs to the town rather
than to the village. In the south, the Vannan, like the Dhobi of Hindustan,
have a subdivision which will wash the clothes of the lowest classes.
In Malabar only the women of the caste do washing and the men work
as tailors. The Nayar have a caste of washermen to themselves, under
the title of Veluttedan, or Vannattan, who often describes himself, at
the Census and otherwise, as belonging to the tribe of his emi^loyers. The
Kanarese washerman is the Agasa. In the Telugu country, the Cakala
have a subdivision which occupies itself exclusively with dyeing, and holds
itself superior to the rest. It seems, indeed, to be connected with the
Velama caste of agriculturists. In the Panjab there is a similar connection
between the Dhobi and the dyer, and in some of the north-central districts
of the Province the two castes are returned impartially by either trade.

§ 45- Fishing, Boating and Porter castes (6,825,400). Of the large
and numerous castes which look back to fishing as their traditional oc-
cupation comparatively few now exercise that calling as their principal
means of subsistence, and these are localised, of course, on the coast and
along the larger rivers. Those communities which have abandoned fishing
have become, generally speaking, separate subcastes, which regard them-
selves as superior in position to those who remain faithful to the net. In this
process of refinement, the first stage is usually the restriction of the ancestral
connection with the water to boating and sea-faring. In the many tracts
where fish is not a staple food among the masses and where there is an in-
sufficient opening in the boat and ferry line, the fisher castes took to the
porterage of such burdens as can be conveyed by poles across the shoulder,
such as packages and large jars, or travellers by palki. It is probable that
in the days when the latter mode of communication was the only alter-
native to walking or riding it fell to the bearers to provide the means
of quenching the thirst of their fare in mid journey. At all events, now-
adays, except in South India and the Dekkan, water brought by those
castes or subdivisions which no longer catch fish is accepted without
cavil by the highest classes. As water is the element above all through
which personal contamination can be conveyed, the privileged position
thus conferred upon the castes in question became assured, and the next

68 5- Ethnography.

step forward was the admission of the caste into domestic service in the
house. This was followed by the recognition of the fisher caste as public
cook, to the extent of parching grain and preparing sweetmeats for the
community at large, and selling them in shops. Thus, in the north and
east of India to which the above remarks mainly apply, the fisherman
basis is found in the Bhadbhunja, the Kandu and the Bhatiara, or cook
of the Panjab, all of which, with a few others of similar trade, are now,
for all practical purposes, entirely distinct castes. Elsewhere, the separation
has been ecjually exclusive, though manifested only by subdivision of the
main caste. The Jaliya or IMecho Kaibartta of Bengal, for instance, the
chief fishing community of the coasts of that province, stands lower than
the Haliya, or ploughing division. The Koli, too, of the west coast, is distinct
from the Talabdii, or agricultural section of this caste, and is called Machi,
or fisher, along-side of a separate caste of that name, one of whose main
subdivisions is called Kujf. The Bhoi, again, has two separate sections, the
freshwater fisherman and the porter or servant. The Boya, of Telingana,
which appears to be the nucleus of the caste, is divided into a village
or settled section, which fishes and engages in service and porterage,
and a nomad, or hunting section, living by fowling and the sale of
jungle-produce. The same distinctions are found in some form or other
among the great fishing castes of the Ganges valley, above the Delta. It
seems probable that these all s|)ring from some K61 tribe of the north
Vindhya, which spread from the hills down the rivers. A great number
of the fisherman are returned at the Census under the general title of
Mallah, which, being Arabic, must have been conferred upon them at a
comparatively recent date. Its subdivisions include many who are else-
where returned under what are usually considered to be distinctive caste
titles, such as Tiyar, Malo, KCvat and the like, with their endless
subsections. One of the castes thus split up, the Patni, appears to be
of a north-Gangetic origin, possibly descended from some sub-Himalayan
tribe like the DOm. The Malo, also found principally in north Bihar,
holds an almost equally low position. The Tiyar comes between the Malo
and the Jaliya Kaibartta. The Kevat in Oudh and Bihar, though probably
of the same Vindhyan origin as the Malo and Tiyar, is largely engaged
in cultivation, and takes his stand, accordingly, above the sections of
the caste which carry loads or engage in domestic service, as well as
above those who still live on the river. In the Central Provinces, the
Kcvat has not abandoned the traditional occupation, and is found mainly
along the Mahanadi and its affluents. There is a colony of this caste in
east Bengal, where, however, they do not catch fish but buy up and re-
tail the haul of the Kaibartta, whom they therefore consider their inferiors.
Above the tract occupied by these castes, the Kahar, or Dhimar, is by
far the most important of the group, and with it comes the Jhinvar of
the Panjab, still higher in position. All these are closely connected both by
rank and functions. The latter are numerous and varied. The Kahar or
Jhinvar is a valuable member of the permanent village staff, and receives
his share of the crops. Though low in relative rank he is pure, to the
extent that he can bear water to all, and enter all but the inner pene-
tralia of their houses. Indeed, in parts of Hindustan, one of the subdivi-
sions is called Mahra, because he is allowed inside even the women's
apartments in the execution of his domestic duties. The Kahar is often
a cultivator in the east, but to the west, he fishes, sinks wells, makes

Castes and Caste-Groups. B. The Village community. 69

baskets, carries burdens and above all, provides the water for the re-
freshment of the peasant in the field. He has a special branch of culti-
vation under him, to wit, the growth of water-nuts (trapa bispinosa), in
the village tanks. His wife, too, as has been mentioned above, is, the
midwife of the Jat and Rajput. The l^Iachi is the counterpart of the
Jhinvar in the west of the Panjab and performs the same duties, with
the e.xception of carrj-ing burdens, the shoulder-pole and palki not being
customary in those parts. There is also a keen demand for his services
as village cook, because in the hot weather the village usually gets its
meals from a common kitchen or oven. Down the Indus, however, and
on the west coast, the IMachi is a fisherman only, and the same may be
said of the Mo ha no, a lower caste of the Sindh waters, which is probably
an occupational body.

In the Telugu country, the Boya, mentioned above, is probably akin
to the Irulan, a wild, roving tribe of hunters and haunters of the scrub-
jungle of the lower hills. The more prevalent fishing caste is the Palle,
which is said to be a branch of the great labouring caste of PaUi, further
south and included in it. The latter was once subdivided into the Mina,
or fishing, and the Vana, or settled, clans, but apart from the barrier of
a different language, the dividing line of occupation now leads the field-
worker to repudiate the fisher, and not to eat or intermarry with him.
Another Telugu caste, the Besta, is, like the rest, both fisher and cook,
and some of its members hold land. They are supposed to be connected
with the Karnatic Kabbera, or Ambiga, who, in turn, form a link with
the coast castes of the Mogcr and INIukkuvan, which go to sea, and
the IMugayan, which fish only in the river. There is a similar distinction
between the Tamil caste of the S'embailavan and their subdivision the
S'avajaikkaran, the seafarers being reputed to rank higher than
the freshwater people. The S'embadavan call in the local Brahman, and
the I\Ioger make use of the Havika, but the rest do not trouble the priest
of any community other than their own.

ij 46. Stone, Salt and Lime-workers (2,043,600). These may be taken
as subsidiary to the fishing castes, since in many parts of the country
the latter have been compelled to take to such means of livelihood, whilst
some of the castes specially devoted to these trades are also connected
by descent with the fishers. The Kevat, for instance, in its lower sections,
is merged into the Bind, and the Bind, in turn, touches the Cain and
the Goiirhi, some of whom are returned as sections of the Mallah. The
majority of all these castes, however, are field-labourers, stone-workers
and lime or salpetre makers, in addition to the fishing or boating sections.
Some of the trades have become the attribute of a caste, as the Luniya,
Rehgar, Soregar, originally functional bodies. The Luniya, or Nijniya,
is the nearest to a real caste, but it is not yet organised on the normal
lines. It repudiates, however the Cain, though probably, their origin is
identical. The latter, in the southern parts of the upper Ganges valley,
has but a poor reputation, not entirely undeserved, for frequenting places
of pilgrimage, with the object of cutting the knots in waistcloths which in
India serve the purpose of a pocket. North of this tract, however, the Cain
ranks low, though with untainted reputation. The Bind, too, stands higher
in rank in the west than in Bihar, whether he fishes or labours in the
fields. On the west coast there are two bodies of salt-workers now driven
to other trades. The Kharvi of Gujarat are sailors and tile-turners, ori-

70 5- Ethnography.

ginally belonging apparently to the Kharol or Rchgar of Rajputana,
who still, like the Agria, are in a position to keep up their eponymous
trade, both on the coast and by the Sambhar lake. Further south, the
Patharvat, now a separate caste, is an offshoot, it is thought, from the
Uppara of Kanara, and are stonc-workcrs, the rest of the community
being earth-workers and carriers by bullock; whilst the Uppiliyan and
Kaduppattan, originally of the same trade, have added the profession of
hedge-schoolkceping to their means of subsistence. The Agria, a Rajpu-
tana caste, still finds room for its traditional making of salt along the
Bombay coast, and to a minor extent in south east Panjab and in the
Agra Province, which, according to some, derives its name from the saline
character of the soil. Where this caste is in force it ranks with the lower
grade of cultivators. In some parts the Agria is held to be a subdivision
of the Luniya, but there seems reason to think that it is a distinct
caste. The Cunari, or lime-burner and the SOri-gar saltpetre-maker, on the
other hand, where they are not separate castes, belong to a branch of
the salt-workers. In Bengal, however, the Haiti, which burns shells into
lime, ranks among the impure, though the product of their labours does
not pollute those who make use of it.

§ 47. Toddy-drawers (4,765,400). Between the lower artisans and
the field-labourers may be taken the castes which live by tapping the
palm for its juice, in some parts of India a body of numerical importance.
They occupy but a low position, partly by reason of their origin^
partly again because the toddy they provide is often kept till fermented,
and being thus an intoxicant, is relegated to the impure articles of con-
sumption. This is the case still more markedly with the distilling castes,
which are classed among the urban and dealt with separately. Along the
coasts the coco and palmyra abound, and the date flourishes in Telingana
and the Gangetic valley. It is here, therefore, that these castes are in
greatest strength. In lower Bengal and on the Gujarat coast, though the
material in question is abundant, it is the custom of the cultivators to
tap their own trees or to employ the ordinary field-labourer or lower
village menial to do the work for them. The tree-tapping castes, too, even
where there is the greatest field for their labour, are largely engaged in
cultivation, either as landholders or labourers. The chief caste of this
description in the Ganges valley is the Pasi, a name derived from a noose,
probably in reference to the belt by means of which the palm is climbed,
or, where the caste is addicted to wandering in the jungle for hunting
purposes, from the snare then used. In Oudh, where the Pasi has a bad
reputation, the noose in question used to be identified with that used by
the Thag in strangling their victims. The Pasi is probably of very early
pre-Aryan origin emanating from the Vindhya, and akin to the Arakh and
Khatik castes, now differentiated by occupation. In Bihar it ranks with
the Bind or Cain, already mentioned as low fishing or boating castes,
but in the west, it takes a lower place. The Bhanijari, of the west
coast, which is not to be confused with the Barber caste of Orissa, ad-
heres more closely to its traditional calling, probably because its oppor-
tunities are greater, and the "toddy-habit" is more extensively established
in the tract where it resides. Its members cultivate also to some extent,
since restrictions upon the extraction of toddy were imposed by the govern-
ment. They also distill spirit from forest produce and sugar in the State
distilleries. Further down the coast, the Bhandari is replaced by two similarly

Castes and Caste-Groups. B. The Village community. 71

localised castes following the same trade, the Paik and the Billava.
Both names are derived from the military services rendered to the Tulu
chiefs by the ancestry of the communities in question. The Paik were the
infantry, and on the strength of the tradition, some of them now claim
to be Ksatriya, substituting the sub-title of Namdhar, for that of Haje,
or old, Paik. By some, however, their name is derived from Pal, the spirit
worshipped by tree-tapping castes. There are probably as many cultivators
among them in the present day as tree-tappers. They speak Kanarese,
whereas the Billava, further to the south, are a Tulu caste, and, share,
moreover, the customs of Malabar in religion and ceremonial, employing
their own priests, where the Paik call in the Satani, an upland caste. The
name Billava means archer, corresponding to the Dhanuk a labouring
caste of upper India, the Kandra of Orissa, and the Cavada, a Gurjara
Rajput clan. The south of the Peninsula is occupied by three large tree-
tapping bodies, probably connected with each other in origin. The name
II a van, which is now used to designate one only of the three, was once
applied to all. It means a native of Ceylon, and the Tiyan, who are
sometimes called by it in south IMalabar, also derive their name from
dvipa, an island, and claim to have come from the south. Furthermore,
they address each other by the name of S'enan, which apparently corres-
ponds with S'anan, the tree-tapping caste of the south-east. They are
divided, like the Nayar, into two distinct bodies, the northerners and the
south-lMalabar Tiyan. The northerners are wealthier, better educated and
more enterprising than the others, and have managed to get some of
their community into good posts under the Government. The southerners
are poor, illiterate, and more closely connected with their traditional
employment, with field labour as the alternative. Still further south there
is a smaller body, the Tandan, probably a sub-caste of the Tiyan, but
not intermarrying with them. This caste has the curious custom mentioned
in connection with the Nayar, of prohibiting its women from crossing a
certain river. As those on the south are far better off than their kinsfolk
on the other side, this restriction may have a solid mundane basis. The
third of these castes, the S'anan, is found principally in Tinnevelli and
Madura, though it is spread to some extent over most of the Tamil district.
The title is not found in the early Tamil dictionaries, and in the inscrip-
tions of the loth century the caste is called Iluvan. The name S'anan is
said to be derived from san and nar, signifying a span-long noose, thereby
corresponding to the name of the Pasi of upper India. The caste came
into great prominence in 1899, when it asserted by force its right to enter
the temples of the Maravan caste, on the score of its K.satriya origin, a
title rejected by the rest of the community. The occupation of the caste
is undoubtedly of great antiquity in southern India, and the Kadamba
dynasty of Mysore sprang from one of its subdivisions. Numbers of the
caste, therefore, were employed in its army and afterwards settled as a
semi military peasantry or labouring class upon the land occupied. The
tradition of such an origin, however, has not survived amongst the S'anan,
whose claims are of comparatively recent date. Curiously enough, the only
sympathisers with the claim, outside those who put it forward, are the
Christian converts from the caste. The general position of the S'anan in
society is that of the lower field labourer, just above that of the menial
class. In former years, indeed, it appears that the S'anan, like the weavers,
were prohibited from living within the village site. In the Telugu country

72 5- Ethnography.

and the Coromandel coast the tree-tapping castes are fairly strong. The
1(1 ig a, which is the principal body amongst them, is an offshoot of the
great Balija class, with whom it still sits down to meals. The sejiaration
seems to have taken place on functional considerations, though the I'.liga
eschew si)irituous lifjuor and employ Brahmans of good position. They pay
special homage, however, to the goddess of toddy and intoxicants generally.
It is sometimes returned as Indra, but the derivation of Ii.liga, from the
verb to extract or draw, like that from the climbing-loop in other cases,
seems to indicate the more ai)])ropriate title. The GamaUa, or Gaunclla
caste is also one of the same locality, and has<a subdivision of the name
of idiga. Its position, however, is a little lower, and it ranks with the
petty cultivators or more respectable field labourers. Brahmans are called
in for its ceremonies, except for funerals, which are under the Satani. On
the coast just below Orissa, are two small castes, the Segidi and the
Yata, which arc toddy-drawers by tradition and mainly in practice. The
latter also weaves mats and baskets from the palmyra-leaf, in spite of its
title, which refers to the date-palm. In the other parts of India there is
either not enough occupation for a special caste of this description, or
the work is done, as in the Central Provinces and Rajputana, by the PasT
or similar castes, already mentioned.

§ 48. Field-labourers. (16,158,400). The castes which come under
this heading are but a fraction of these whose members make their living
to a great extent by field-labour. The rapidity with which crops come
to maturity in the tropics and the shortness of the time available for each
harvest produce an urgent pressure upon the labour supply, which is met
by the temporary diversion to the fields of numbers who during the
rest of the year follow quite different occupations. Even the normal demand
is very great. There is to be taken into account the universal prevalence
of agriculture, and the vast numbers of holdings which require more hands
upon them than can be furnished by the occupant's own family. Then,
again, there are some important operations which are not lawful for the
cultivator of high caste, entailing, therefore, the permanent employment
of menial hands for the purpose. These are procured from the village
servile classes, the rest of whom have their own special caste functions.
Thus almost all the lower grades of the rural population contribute a
certain quota of agricultural labour. In former days the system of predial
servitude was widely spread, and whole castes were assigned to certain
families or estates in a district, as on the Malabar coast and amongst Brahman
agriculturists wherever they are found; and though the status of the la-
bourer has been changed under British rule, the practice, on a voluntary
basis, still persists. In some other parts of the country the labourers are
distributed by families, each ascribed to a certain employer, or patron,
from whom they receive special gifts or privileges beyond the mere re-
muneration of their labour. Finally, there is the constant transition of the
landless labourer, by thrift and industry, to the position of petty landholder,
not unfrec|uently accompanied, after an interval, by the severance of this
class from the less fortunate of the body in which it was born. Thus,
whilst the upper edge of the group overlaps that of the humbler landed
classes, the lower is merged in the general body of the impure or servile
castes at the bottom of the village community. In the group now under
consideration an attempt is made to include only the upper stratum ot
the castes traditionally dedicated to field labour, and to deal with the rest

Castes and Caste-Groups. B. The Village community. 73

separately. It must be admitted, however, that it is almost impossible, in
view of the different standards in force, to draw the line accurately.

Amongst the Dhanuk, for instance, a caste spread over the Jamna
valley as well as north Bihar, the position is apparently higher in the
latter tract, and might fairly entitle the caste to be ranked with the minor
landed classes. This is not the case, however, elsewhere, and the fact
that the most esteemed subdivision in Bihar is that in domestic service,
and to a great extent born on the premises of the employer, seems to
indicate that the peasant section also is one of "new men". From the name
of the caste, which means Archer, like that of some of the corresponding
castes in the Dravidian country, it may be conjectured that the Dhanuk
were once a local militia, reduced in circumstances, for in the Agra pro-
vince, they are the village trumpeters, and their wives share with those
of the Barber the office of midwife. In Gujarat there is a similar case,
that of the Dhodia or Dhundia, a tribe of KiJl origin left on the plains,
which is rapidly passing from the labourer into the occupant, whilst the
Dubla, its congener, who fell at an early stage into the hands of the
cultivating Brahman, is still in a state of practical servitude on the farms
of the latter. It is true that in the great "cotton years" of 1S63— 66,
the Dubla took to free labour, but, for the most part, they found it more
advantageous to revert to what is now called hereditary service. Re-
verting to upper India, the Arakh, a small offshoot of the Pasi, is
undoubtedly a fallen caste, for it held a tract of the valley against the
Rajputs, and was only subdued by the Muslim in the I4'i' century. It still
ranks above the other Pasi, but labours for its bread or acts as village
watchman. In the west of Bengal are found two castes of Kul origin,
but long settled in the plains as landless labourers, a few holding land.
The Bagdi probably rank a little above the Baurf, as being more par-
ticular in their diet. They are carriers of burdens, hewers of wood, and
workers in the indigo fields. Both castes admit into their community
members of higher castes who are in need of such a refuge, but no
recruits are accepted from below. They are described as being just "on
the outskirts of Brahmanism". In Bihar and the east of Oudh are the
Rajvar and Musahar, low castes of labourers of Kdl descent, or, at
least, belonging to the dark races of the Central Belt. The Rajvar stand
the higher of the two, and employ degraded Brahmans for their cere-
monies. They have retained a good deal of their tribal organisation but
have settled down to cultivation and labour. Some of them have acquired
holdings, as tenants, but have not yet risen above this grade. According
to their own account, they belong to the same stock as the Musahar, but
stand higher. There has been a good deal of controversy as to the latter
caste. The name is said to mean rat-eater, a habit the caste still retains,
and this is one of the reasons why the Rajvar, who does not indulge in
this diet, will have no communion with his kinsman. That the two are
both pre-Aryan is certain, but whether the descent is from the K6I through
the Bhuiya, or Dravidian through the Ceru, is undecided by the authorities
on the subject. The JNIusahar has not yet been organised on ordinary
Brahmanic lines, and retains much of its primitive form of worship along
with its tribal subdivisions. Brahmans are occasionally called in, but most
of the ceremonial is carried (m without sacerdotal aid. The JMusahar are
divided, like the Boya and other tribes of their calling, into two sections,
one settled in villages, carrj-ing loads and doing fieldwork, the other

74 5- Ethnography.

haunting the jungles and collecting wild produce, which they bring for
sale into the villages. One of the reasons given in Bihar for employing
men of this caste to watch crops in the fields is worth noting, viz that
the Musahar is alone able to keep off the older gods, who have been driven
away by the plough and resent the intrusion of the alien peasantry. West
of the Musahar is found the Bhar, now holding a higher rank than his
neighbour, but bearing in his physical appearance manifest signs of his
descent from a similar dark race. The Bhar is said to have once held
the land on which he now labours, but was ousted by the Rajputs when
they in turn fled before the Muslim. As the tribe has no tradition of
migration, it is probable that it was formerly in a better position than
now, but it must always have been of unsettled habits, as even now its
favourite occupation is breaking up fresh land; and when a village area
has once been brought fully into cultivation, the Bhar is inclined to leave
it for the nearest virgin soil. The Bhar of western Bengal seems to be of
higher position, and employs Brahmans where his northern namesake uses
no priest at all. The latter, too, retains the rites customary among the
Kori and Camar, and owns no connection with the others down the river.
In Rajputana there is a small caste, the Dhakar, which seems to be of
fairly good position, and is employed upon the estates of Rajputs; but the
field labour generally, both here and in the Panjab, has fallen into the
hands of the leather-working and impure castes. It is the same, for the
most part, south of the Vindhya, as far as the Dravidian country, and
some sections of the Koli are the only castes which can be said to be
specially field labourers of a superior grade. The contamination which
follows upon the use of the same implement, drinking out of the same
vessel or of the same water, or smoking the same hukkah, is avoided,
of course, by a strict demarcation of the various operations in the field,
by the use of differently shaped lotahs, and by denoting the pipe of each
caste by a differently-coloured rag tied round it.

§ 49. Dravidian Labouring castes. In the south of India the landless
labouring classes are particularly strong in number and assertiveness, and
their relative positions are hard to define and must be treated as doubtful
pending the results of the investigations of the Ethnographic Survey. It
is advisable, therefore, to deal with them apart from the rest. There is
apparently some reason for believing them all to be of one origin, but
superimposed at different times one upon the other by various waves of
conquest or migration. Their position has thus varied more than that of
the corresponding helot tribes of the region absorbed by foreigners from
beyond the north-west of India. The title Paraiyan, for instance, is not
found in the standard Tamil dictionary of the ii''> century, but the caste
now so called is referred to in contemporary records under the name of
Pulayan, still used of the corresponding community on the Malabar coast.
Some weight may also be attached to the similarity of these two names
with those of the PaHi and PaUan, labouring castes of the south Tamil
country. The Holar or Holeya of the Karnatic, too, appears to belong to
the same group, as in Kanarese the Tamil P becomes H. The PalU, to
whom the name of Vanniyan w-as given by the Brahmans. were once a
dominant tribe under the Pallava dynasty, but were reduced to predial
servitude when the Vellalan entered their country. They are now mainly
agricultural labourers, though some have acquired land of their own and
others engage in trade. They occasionally call in Brahmans for their rites.

Castes and Caste-Groups. B. The Village community. 75

but their customs and rules are for the most part purely Dravidian. On
the score of their former position, they have of late put forward the claim
to be considered K.satriya, and don the sacred thread, conduct which
brings them into collision with both priest and peasant. It is said that in
the Right and Left-hand distribution of castes in the Tamil country, the
men of the Palji go to one side and the women to the other, conjugal
relations being suspended whilst the facti(ms are in active opposition and
resumed when peace is temporarily restored. The PaUan, in spite of the
similarity of the name, own to no connection with the PaMi in the present
day, and occupy a tract to the south of the latter. They arc lower in
rank and rarely engage in pursuits other than field labour. The names
of their subdivisions, however, indicate that they may have belonged to
the great Kurumban tribe and thus have an ancestral connection with the
Pallava and therefore with the Palli. They follow the regular demonolatrous
worship of the older Dravidians, and if they use priests from outside, they
call in the Yelluva, a low caste ministrant. The Pulayan, mentioned above,
is a labouring caste of north Malabar, called Ccruman in the southern
portion of that tract. They have a traditiim of better, even dominant, days,
before the Nayar enslaved them on their estates. One of the relics of their
servile condition is the practice of still bringing their children to be
named by their employer. They use their own priests in the propitiation
of the evilly-disposed goddesses they worship. In a good many respects
they follow the customs of the Nayar, such as inheritance through the
female line in the north and through the male in the south. The title of
Ceruman denotes, according to their tradition, an origin in the Cera country.
There remains the great community of village menials of a type more
pronouncedly impure than the castes mentioned above. These rank above
the tanners and leather workers generally, and above the scavenger,
whether a separate caste or, as in the greater part of the south, a sub-
division of the main body. The best known section of this group is the
Paraiyan or Pariah, of the Tamil country. In treating of it it is advisable
at the outset to get rid of the notion set on foot by the Abbe Raynal,
that the Pariah is an "outcaste", or that there exists such a thing as an
outcaste anywhere in India. Every community has its place, disputed though
it may be, in the social hierarchy of Brahmanism, and there is no caste
but will unhesitatingly designate some other as ranking below it. Ethno-
graphic inquiry, therefore, past and present, has never yet succeeded in
touching the bottom, or in finding a waif for whom no recognised place
exists within the fold, albeit without the village. Possibly, in the course
of time, public opinion may crystallise round one of the nomad castes,
who know nothing of their past, and recruit and eat as circumstances
dictate. Meanwhile, the scavenger fills this situation in the village life with
which this review is at present concerned. Now, the Paraiyan is a caste
the position of which is at all events clearly defined, and it has a past
which it cherishes. Low as he is, excluded from everyday communion
with those above him, "le morne Chandal" will no more admit the pol-
luting presence of a Brahman into his hamlet than the latter will allow
the Paraiyan's shadow to fall upon his water-pot. Some of the most
celebrated and exclusive temples are thrown open to the Paraiyan on
certain days of the year, and for the time he lords it over the Brahman.
At certain festivals again, especially those connected with S'iva or a local
goddess, it is one of this caste who takes his seat alongside of the image

76 5- Ethnocraphv.

in the procession, or ties the symbolic marriage-thread round its neck.
Until recently, when the custom began to wane, even the Brahman, in a
few tracts, had to obtain the formal consent of the Paraiyan to a marriage
in his household, and similar acts have been mentioned in connection with
the rites of castes dealt with in a preceding paragraph. In another direction,
certain low but responsible offices on the village staff must be filled by
Paraiyan, and when there is a dispute about a boundary, it is a Paraiyan,
or," in other parts of India, a member of the corresponding caste, who has
to walk the line with a pot of water, his own son, or a clod of his native
earth, on his head. All this tends, of course,- to show that the caste was
once a most important element in the population, older on the soil, in
closer communion with the genius loci, and influential beyond the con-
ception of those who only know it in its condition today. As before pointed
out, its present name is comparatively modern, and in the earliest records
available, before even the I'ulayan are mentioned, the caste which, like
the Paraiyan of to day, was excluded from the villages, was called Eyinan,
and credited with the possession of hillforts and considerable power, on
the lines of the Dasyu of the Sukta period. The sub-castes of the Paraiyan,
which are very numerous, indicate the practice of most of the more re-
putable handicrafts, but the general tradition among the modern Paraiyan
is that the caste was formerly a weaving one by calling, and in an in-
scription of the nth century, probably the earliest in which the name
Paraiyan is used, it is subdivided into the weaving and the ploughing
sections. Some have derived the name from parai, a drum, and a section
does, indeed, act as the drummers of the Right-hand. On the other hand,
their great rivals, the leather-workers, blow the trumpet for the Left,
without being named after their performance on that blatant instrument.
In the Karnatic, the Holeya occupy almost the same position, except
that they are not, of course, affiliated to any factional distribution of other
castes, nor do they weave to any great extent. A good many of them have,
however, joined the Lihgayats, in which community weavers abound, some
of them holding but a low position, attributable probably to their origin
amongst such classes as the Holeya, and entailing, at all events, the
establishment of a special section for their reception. In the Telugu country,
the place of the Paraiyan is taken by the Mala class, the name of which
resembles that of the Mahar of the Dekkan, which performs the same
offices. In the case of the latter, however, the weaving branch has split
off into an entirely separate body, whereas in the east it seems to remain
as a subdivision. All these Dravidian labouring castes employ barbers,
washermen and generally priests, of their own community. Ethnologically,
the group presents features of very great interest and importance in re-
spect to its origin and history, and much remains to be done in sifting
the different strata of a people of whom so little is known in comparison
with what has been ascertained concerning the servile classes in upper
India. Not that there is any lack of theory, conjecture and analogy.

Two castes of western India may be here mentioned, which arc de-
dicated generally to the same functions as most of the castes just reviewed.
One of them, indeed, the Mahar of the Dekkan, is probably allied, as
stated above, to the INIala of Telingana. The distinction, however, in these
tracts between the depressed castes and the rest of the village com-
munity is more definite than in the south, partly, no doubt, because racial
differences are greater or have been less obscured by time. The Mahar,

Castes and Caste-Groups. B. The Village community. 77

for instance, belongs to a far earlier race than the Maratha peasantry,
and enjoys a notable prestige amongst them for knowledge of the boun-
daries, and for influence with the goddesses of cholera and small-pox. The
caste, too, has its own priests, but near the larger towns as often or not
a Dcs'asth or local Brahman is called in. This is, however, a modern
practice, introduced since the labour market on railways and large public
works brought grist to the INIahar mill. Formerly, and perhaps even now
in some tracts, the Mahar had to wait for a ceremony amongst the higher
castes, and then bring his own party up to just beyond the prohibited
range, so that the sacred texts could be heard, with the fiction of the
impure listener being out of earshot. The Mahar is as a rule, a labourer,
and those who take to trades separate themselves from their fellows. The
caste, like the Paraiyan, holds a low but important and useful place in the
village staff, and receives shares of all the main crops, and, in some places,
a considerable piece of the land. The Dhci.l caste of Gujarat, on the
other hand, is not one of the recognised community of the village, except
in the south, and even there he is not regarded as one of the old stock,
and has no special knowledge of the boundaries or of the idiosyncracies
of the local gods. In fact, he is apparently what he claims to be, an
immigrant against his will from Rajputana, though the tradition of the
movement is no longer definitely retained. In the north of the province,
the menial work of the village is done chiefly by the Bharigi, a lower
caste, and the Dhed was until recently, a weaver of coarse cotton goods.
When factories were established in Bombay and the chief towns of Gujarat
the Dhed lost much of his custom, and took to working under the new
regime at the machine-made article, whilst others took to day labour.
North of the Narbada, the families of this caste are often found attached
to the estates of the larger Kanbi or Rajput landholders, by whom they
are supported. In the south a special sub-caste has been formed of those
who have taken to domestic service with Europeans, here again following
the same lines as the Paraiyan. Either on account of this adaptability or
because of the thrift displayed by the caste in its various callings, the Dhed
is credited in a local proverb with having profited above others by British
rule, and to have waxed fat and kicked accordingly against his Brahmanic
betters. Though the caste employs only low caste priests it is credited with
great orthodoxy and assiduity in its religious duties, as well as with strict-
ness in the observance of the rules of the caste, enforced by local councils.
§ 50. Leather-workers (15,028,300). This group, as was stated above,
cannot be well distinguished from that which precedes it. It is the function
of all the impure castes to deal with dead cattle, even if it be only to
skin and to drag the carcasses away for burial. But there are grades and
privileges involved. Some touch no bodies but those of the cloven-footed
animal ; others draw the line at cattle, and leave sheep and goats to
their inferiors. Usually the hide is the perquisite of the menial, who,
moreover, is not forbidden to indulge in the flesh after flaying. Indeed,
when the market for leather is brisk, or when dissension is rife between
the peasantry and the village menials, mortality amongst the cattle is apt
to increase materially, and sometimes with a suddenness which attracts
the judicial attention of the local authorities, and leads to the discovery
in the thatch of the servile hamlet of the materials for an extensive study
of rural toxicology. But though the castes in question remove the hides,
it is only special sections of them which tan or curry them, and these,

jg 5. Ethnography.

except in the north, arc generally split off into a separate caste. Further-
more, the families which take exclusively to leather-work as their pro-
fession beyond the simple requirements of the cart, plough or water-lift,
usually rise to a position superior to that of the tanner or currier, and
ultimately, especially in towns, hold themselves aloof from the rest. On
the other hand, where the caste furnishes virtually the whole labour supply
of the village, the tanning branch sinks below those which only labour
in the fields. In the latter capacity, the caste has to do whatever they are
bid by the peasantry — within, of course, the strict bounds of tradition.
They may never, however, take up their residence in the village or pass
anything directly from their own hand to that of one of higher caste. It
is a noteworthy fact that with centuries of such degradation piled upon
them, the women of this class should be renowned for their good looks;
so much so, that special arrangements seem to have been thought neces-
sary by the Brahmanic organisers of society to meet the results of intrigues
and illicit connections between them and men of the upper classes. To
this day men turned out of their caste on this account find refuge in
some recognised mixed body, whilst the offspring of such mesalliances
go to form the "fair-skinned Camar", the subject of more than one pro-
verbial admonition on the country side. There is the possibility, of course,
that in the very north of India some of the helot classes may be
descended from early foreign races who were overwhelmed by subsequent
invaders and reduced to servitude, but throughout the rest of the country
these classes are now generally held to represent the Dasyu or darker
tribes, displaced by the Arya and Scythian invader north of the Vindhya,
and by similar movements amongst Dravidian races and others, in the
south and the great delta of the east.

The great Camar caste is found all over the country except in the
south, but in the tract where it is most numerous, between the east Panjab
and Bihar, it is not exclusively a leather-working caste as its name de-
notes. It supplies, as just pointed out, the main body of field labour, and
receives its share of the harvest like the other village menials on the
establishment. In this capacity, the Camar community is generally organised
into distinct sections, irrespective of social subdivisions. Some work for
individual patrons, but more often each is assigned to a certain association
of landholders. The development of the leather industries upon European
lines in some of the large towns of the north, such as Cawnpore and
Agra, has attracted a large number of Camar away from their native
haunts. Indeed, the demand for labourers along the railways and in the
chief commercial centres of upper India is said to have had the effect
of depleting to a considerable extent the supply available for the village
field operations, and the Camar, like the Dhcd of Gujarat, leaves home
when he pleases, and returns with a full pocket and something of a
"swelled head". In parts of Rajputana and the southern Panjab, the Camar
docs the coarse weaving undertaken further cast by the Kori. The caste
is subdivided minutely by function, locality and traditions as to origin,
into endless endogamous sections, in a recognised order of precedence,
and all under the regulation of a caste-Council which is said to be strict in
its enforcement of ceremonial rules. In the central and eastern Panjab a good
many of the Camar are Sikhs by religion, though of course they occupy a
position different from that of the Jat. Comparatively few seem, from the
Census, to have embraced Islam, but this is due to the use of the title of

Castes and Caste-Groups. B. The Village community. 79

MOci by converts, especially in the west of the Panjab, where they are nearly
ail Muslim. In other parts of India, the MOcf is the subdivision, generally en-
tirely distinct, which is engaged in shoemaking, usually in the larger towns.
Even in the west Panjab the Camar or Moci do not perform the same duties
in the village as the Camar of the east, but only do the leather-work and
tanning, thereby taking a higher position than their agricultural fellow. The
Camar of other Provinces is a Brahmanist in his faith, of much the same
order as the lower masses of the population of the locality. In some parts he
gets Brahmans of a low grade to serve him, but, as a rule, they are only
called in to nominate the most auspicious day for important domestic cere-
monies. By reason of the connection of the caste with the exuviae of dead
cattle, the Camar is held to be lower in rank than even the Brahmanised
section of a converted forest tribe which has abandoned the cruder elements
of its daily diet. It does not appear, however, that this was always the case,
as leather entered into the clothing of the early Vedic communities long be-
fore they could have reduced the Dasyu to servitude, so that the task of
tanning and preparation must have been performed by members of their
own race. The degrading character of the occupation, therefore, may have
been imputed to it by the Brahmanic censors of the new regime when it
was established upon priestly initiative at a later date.

In the lower Himalayan valleys of the Panjab there is the Mcgh caste,
who perform much the same duties as the Camar of the plains, but are
rather higher in social esteem because they are largely weavers, and leave
the dirtier offices of the village to lower castes, such as the Koli and
Dagi. The latter do the leather work in some parts, but elsewhere they
put it on to the Koli or Canal. All are of about the same class as the
Camar, some even being subdivided under that title, and represent the
earlier tribes of the locality, reduced to servitude by the later comers
from the south or west. They resemble the lowest castes of the plains,
too, in acting as pipers and drummers at village processions. South and
west of the Vindhya, the caste is still known by the names of Cambhar,
or Khalpo, but is quite unconnected with the northern communities of the
former name. The leather work, too, is detached, more or less, from the
menial offices, and is not intimately bound up with the village staff.

In the Dekkan and Telugu country, the Camar gives place to the
Maiig or Madiga, both of which names are apparently derived from Matafigf,
the caste goddess, a synonym of Kali. The JMadiga takes a prominent
part in the festivals of the S'akti worshippers, probably of Dravidian origin
incorporated into the Brahmanic pantheon as circumstances demanded.
From this as well as from the part it plays in the marriage ceremonies
of some of the higher castes, it may be inferred that the caste is one of
the earliest of the uplands, and thus more likely to propitiate the local
gods than the more reputable but more recent arrivals now in occupation.
Both Maiig and Madiga employ their own priests, Garuda or Dasari. Where
the Maiig is found alongside of the Mahar in the Dekkan there is always
rivalry and occasionally strife, but the Mahar takes precedence of the
other in the village. In the Tamil country the principal leather-working
caste is the S'akkiliyan, vulgarised by Europeans into Chuckler. It is
an immigrant body, as several of its subdivisions bear Telugu or Kanarese
titles, and many of its members still use those vernaculars. It may be
added, too, that its name does not occur in any of the older inscriptions
in Tamil. It is probably, therefore, an offshoot of the Madiga, moved south.

So 5. Ethnography.

importing with it its traditional rivalry with the village serf, for there is
constant bickering between the S'akkiliyan and the Paraiyan, public opinion
being in favour of the labourer, as in the Dekkan. It may also be noted
that the leather-workers are here, as in the north, remarkable for the
beauty of their women, and in those stages of Sakti worship at which
the presence of a living representative of the Female Energy is necessary,
a S'akkiliyan girl is always selected for the part.

It is only the simpler leather work, as was mentioned above, that is
done by the village Camar, and though he can cobble shoes, he does not
generally make any but the roughest kinds. The MOci takes up the higher
branches of the craft, but in Bengal, as in the west Panjab, this caste
does a good deal of the village labour, and in the former tract his shoes
are said to be inferior to those of the Camar of Bihar. In Rajputana the
Bambhi seems to be the shocmaking branch of the latter, and in 1S91 some
207,000 of them were returned, but as in 1901 they were reduced to lioo,
it is probable that the rest are included in the main Camar caste. In several
parts of India, the Moci of the towns are divided into functional sub-castes,
such as that of saddlers, embroiderers of saddle-cloths, makers of leather
buckets for ghl (clarified butter), of spangles, shields and scabbards, rising
in rank as their calling entails greater skill or more costly materials,
always tending towards endogamy within the craft.

§ 51. Watchmen (3.639,900). There are few countries, possibly none,
in which the old counsel to set a thief to catch a thief has been more
widely and conscientiously put into practice than in India. In the case
of more than one of the castes already passed under review it has been
pointed out that a portion of the community in question was avowedly
detached for night work of one sort in order to counteract the enterprise
of its comrades in simultaneous operations of another. In several of the
older lists of the castes of a locality, too, there may be found opposite
a title, the terse description, "Thieves and watchmen". The combination
is obviously appropriate in tracts interspersed with hills and forests, or
containing the broken ground, frequent in India, in which the facilities
of both functionaries for evading observation are united: or, again, where
tribes of hunting and fowling propensities have settled down to village
life. But even in the open and well-cultivated plains the need of a night-
watch over cattle, grain and other movable property is generally recognised,
although the underlying notion of blackmail may be absent. In the latter
case, however, the duty is performed by a local caste in which it is not
the traditional or even the principal mode of getting a living. It tends,
however, like all else in India, to become hereditary in the families which
take to it, and, if associated with a recognised dole out of the harvest,
to be ultimately crystallised into a sub-caste. This seems to have been
the case with the Dhanuk of the Ganges valley, though the branch of
the caste which has found its way into the eastern Panjab is treated as
criminal without the saving grace of occasional watchmanship. The Mahar
of the Dekkan, again, has recognised subdivisions of watchmen and the
guardians of the village gate. There are also castes which are traditionally
watchmen without any association with the predatory classes. Among these
are the Barvala and Batval, of the lower Himalayan valleys of the Panjab,
who, though chiefly watchmen and messengers, also perform many of the
menial offices which in the plains are left to the Camar, but draw the
line at dealing with skins and leather. They are not allowed however.

Castes and Caste-Groups. B. The Village community. 8i

to reside within the village site, and in this respect are on the level of
the Mahar. The Ghatval of Bihar, again, has become a separate caste
in consequence of its having appropriated to itself the guardianship of
the low passes through the hills, and has a share in the general name
of JNIallah. But it is most probably an offshoot of that wide-spread and
incoherent tribe known as the Bhuiya. The Kandra of Orissa derive
their name, like the Dhanuk, from their prowess in archery, and in former
days constituted a local militia in conjunction with the Panka. They
are now watchmen and labourers, keeping up much of their old religion
and customs, but cmjjloying Brahmans on occasions. In the Dravidian
country, the Ambalakkaran of the south-ca.stern Tamil districts, have risen
by the adoption of Brahmanic rules from a hunting caste to an cstalilished
village position as watchmen and cultivators. Their kinsfolk, the Mut-
tiriyan, are said to have passed through a militia stage before settling
down to the guardianship of the village. They are affiliated by some to
the ;\Iutraca, a larger caste once no doubt the guards of the frontier
of the Vijayanagara dominions, and it is possible that the military traditions
of the Muttiriyan are due to this relationship. The Wutraca, however, are
from the Telugu country, and the connection therefore may be no more
than is suggested by similarity of name.

There remain the castes which are constituted watchmen more from
apprehension than from an a priori confidence in their efficiency. Amongst
these maybe counted the Khangar of Bundclkhand, now numerically insigni-
ficant, and subdivided into a cultivating and respectable section, and one
which furnishes watchmen and labourers to the villages. It is no doubt one of
the early Vindhyan tribes a portion of which has been Brahmanised by enlist-
ment into local forces and contact with the Rajputs by whom the tribe was
dispossessed of its hill-strongholds. The upper section has no social inter-
course with the watchmen. The latter retains its old customs and religion,
does not employ Brahmans, and, although not one of the regular criminal
tribes, is sufficiently prone to petty theft and burglary to make its enlist-
ment as Kotval or watchman, advisable. In some cases it is returned at the
Census under this name, but it is totally unconnected with the watchman
caste of Bardvan, in Bengal, or that of the west, which is Bhil, or of the
Central Provinces, which is Gond. A more important community of this
class is the i\Iina of Rajputana, to which reference was made in connection
with the Meo, the Muslim and more settled branch of the same tribe.
The Mina are spread all over the east and north of Rajputana, and were
formerly the rulers of a considerable portion of the present state of Jaipur,
if not of Alvar and Bhartpur also. Even now, they occupy a dominant
position amongst the agriculturists of the east, and in Jaipur, a section
is employed as the special guardians of the palace and State treasure.
It used to be the custom, moreover, for a Mina to complete the enthrone-
ment ceremonial of the Chief of Jaipur by affixing upon his forehead the
mark of his caste, just as in Mevad, the Chief has to undergo the same
operation at the hands of a Bhil, in token of the acquiescence of the
former owners of the soil in the new order of things. There is no doubt
that the Mina are of early and pre-.Aryan origin, though a section of
them has been impregnated by Rajput blood to an extent w'hich encourages
them to claim to belong to that order. Of the two sections into which the
tribe is divided, the Caukidari, or watchmen, used to be the terror of
Central India, and carried its raids far south of the Vindhya. As it still

IndoAryan Restarch. IL j. 6

82 5- Ethnography.

exercises its traditional functions of guarding the villages, it considers
itself higher in rank than the other sub-division, the Zamindari, which has
settled down to cultivation, and it used to take its brides from the latter
without returning them. Now, however, the cultivator has advanced in
prosperity and refuses to recognise the older section either as its superior
or even as its equal. In this it was supported by a former Chief of Alvar,
who did his best to sever the more reputable of his subjects from the
contaminating influence of their turbulent fellow-tribesmen. In the south
of Rajputana the Mina hold a lower position than up north, and in
Marvad some rank as village menials of. the impure grade. In the
neighbourhood of the hill tracts they are also hunters and fowlers, and
everywhere their reputation is the basis of their employment on the vil-
lage staff. Almost the same can be said of the Bhil, who, in Gujarat,
serves as watchman, under the sub-title of Vasavo, a name applied to
his tribe in the western Satpura. In Bihar and along the Ganges as far
up as Mirzapur, the large caste of the Dosadh undertakes the duties of
watchman. This community is very mixed. It has undoubtedly a strong
strain of Mongoloidic blood, but it is peculiar in the extent of its formal
recognition of members of higher castes who seek admission to its
ranks. It employs degraded Brahmans for ordinary purposes, but at the
chief festival of the caste, that in honour of Rahu, the demon of eclipse,
one of its own number officiates. The Dosadh used to furnish many re-
cruits to the Muslim armies of Bengal, and it is said that a considerable
proportion of Clive's army at Plassey was composed of this caste. Now,
however, the Dosadh has but a poor reputation for industry-, whilst it is much
addicted to crimes against property, entailing its employment as watchmen.
The rest of the caste get their living by porterage and day labour. The Mai
of western Bengal is largely engaged to watch crops and villages, as many
of its sections are thieves and wandering pilferers. It belongs to a large and
widely-spread Dravidian tribe now divided into numerous separate castes.
A similar caste to the Dosadh is found in the Bcrad, or Bcdar,
'fearless ones", of the south Dekkan. These were originally hunters and
fowlers of the Karnatic, and were formed into militia by the Muslim Chiefs
of Mysore and Haidarabad, in which capacity they served till a compara-
tively recent period. They are now watchmen and petty cultivators. Their
faith is Brahmanic, of the semi-Dravidian type, and they employ the Sa-
tani caste as their priests. Possibly they come of the same stock as the
Buya, one section of which pursues the same calling, or the Vctjan of the
Tamil country, who are still hunters and in the jungle phase of existence.
In the Maratha country, especially near the Sahyadri range, the place of the
Berad is taken by a kindred tribe, also from the south, known as the
Ramosi, a title which is said to represent the Marathi Ranvasi, or
forestdweller. They address each other, however, as Boyali, indicating
Telingana parentage. They stand higher than the Bedar, and employ by
preference, the Jangam priests of the Liiigayat, with a Gosai for their re-
ligious and moral instructor. According to the caste reputation, the func-
tions of this individual are more necessary than effective. By the age of
seven, the Ramosi boy must have stolen something or he is disgraced.
If caught and convicted, the halo thereby acquired renders him a prize
in the marriage market for which an unusually high dowry has to be of-
fered. Another peculiar tenet of this caste is that meat is not to be eaten
unless it has been killed by a Muslim.

Castes and Caste-Groups. B. The Village community. 83

§ 52. Scavenging castes (3,647,700). This group includes the lowest
of at all events the village castes of India, whatever may be their position
relatively to the immoral and foul-feeding nomad. Yet even here there
are gradations of rank duly recognised within the community though not
affecting its intercourse with the outside public. For this reason, perhaps,
the Bhangi or Mihtar caste of the upper Gangetic region is subdivided
to an unusual extent, and the main endogamous sub-castes are strict in
regard to the limitation of their respective functions. Judging from the
nomenclature of the subdivisions it may be inferred that the caste was
originally formed out of a number of local tribes, reduced or compelled
to have recourse to occupations repudiated by the community to whom
they were subject. Some of these sub-castes draw the line at carrying
loads and playing pipes and drums; others have become watchmen, cane-
workers, domestic servants, sweepers of roads, and plasterers of walls
with cowdung. A section which keeps pigs, again, ranks below all but
those who remove night-soil, and amongst these last, those who serve
private houses hold no intercourse with those employed on public latrines.
It may be borne in mind that these latter functions are confined to towns,
except where the women of the household are strictly secluded. Else-
where, the custom of the country renders their offices unnecessary. The
great differences in the physical appearance of sections of the caste do
not indicate a different origin of the respective communities, but a varied
recruitment from higher castes of "broken men"; and, also, the impregnation
of the sections undertaking domestic service with the blood of their em-
ployers through illegitimate connections, the Mihtaranf sharing the repu-
tation of the Camari for good looks. She is also called in, like the wives
of several of the low castes, to perform duties connected with childbirth
which no higher class will undertake. In the west, where there is no
question of a lower caste, the Bhangi will handle a corpse, kill a stray
dog, and act as hangman. Further east, he finds that these functions can
be thrust upon the Dom, a tribe of probably quite as early origin, but
later enslavement to Brahmanic supremacy. In the Central Panjab the
Cuhra does much the same work that the Camar does where the latter
is in full strength, and resents the title of Bhangi. In the west of the
province the ^Muslim sweeper known as Kutanan or Musalli, digs graves
but will not touch night-soil. Further to the south, the Ciihra is called
Jat like many other menial castes. In the east, the caste is a recognised
member of the village staff and belongs to the Bhangi community of the
Gangetic region. Amongst other duties may be mentioned one of great
importance in a land where fuel is scarce, that is, the collection, drying
and storing of cowdung for burning. The sweeper, too, is the only caste
which will convey the tidings of a death to those whom it may concern.
In the Sikh tracts many Ciihra have joined that faith and after conversion
continue to perform only the less offensive parts of their traditional du-
ties. One of their subdivisions, the Rangreta, has risen in position by
taking to leather work exclusively. The Mazhabi, or Mazbi, as the Sikh
Cuhra is called, makes a capital soldier, but has to be brigaded in se-
parate regiments, as the other Sikhs, with their eye on the traditional
calling, refuse to associate with the convert, even in religious ceremonies.
Occasionally the Sikh intermarries with the Lai Begi, or Brahmanic Bhangi.
In north Gujarat, the Hhahgio is one of the principal village menials, and
does most of the unskilled labour. In spite of the Rajput titles of the sub-

84 5- Ethnography.

castes, this community is one of long settlement on the land there. It is
the Bhangiu, for instance who points out the boundaries; the sight of
one of this caste carrying his basket brings luck for the day, and before
crossing the Mahl river in a flood, the blessing of a Bhangio tends to a
safe passage. In this part of the country, as on the Ganges, the Bhangi
is strict in his religious observances, but is only allowed, of course, to
worship from the outside court of the temples. As in the north, too, this
caste has the provision and control of the village music at times of festival.
In Bengal and Assam the chief castes of sweepers are the Bhuinmali
and the Harl, or Haddl. Probably both are of the same stock, a Kol
or Deltaic tribe of early settlement. The Bhuinmali is found in the north
and east of the province, the Hari in the west and centre, and the Haddi
in south Orissa. Both arc subdivided into functional sub-castes which do
not intermarry. Musicians and porters stand highest, and often take to
cultivation. The ^lihtar, borrowing its name from upper India, is the
lowest section, and the only one which touches night-soil. Between these
come sections working in cane, tapping palms and carrying torches at
weddings. One section has taken to private service. The smearing of
wet cowdung upon walls is a frequent occupation of the Bhuinmali, but
they can only touch the outer walls, and except this caste none will
touch the wall of another owner, though each householder does the steps
and inner walls of his own dwelling. The Hari has preserved much of
the non-.\ryan customs of his original tribe in regard to marriage, and
is singular amongst the widow-marrying classes of India in prohibiting
instead of encouraging, the marriage of the widow to the younger brother
of her late husband. The caste does not, as a rule, call in Brahmans,
but the practice of making use of them is spreading round Calcutta, though
the Brahmans in (|uestion are put out of communion by their fellows.

As the Dravidian country is approached the village scavenging is
more and more done by some of the menial castes mentioned in a preceding
paragraph, such as the Paraiyan or Mala. It will probably be found that
as elsewhere endogamous sub-castes are being formed, separating the
sweeping and labouring families from those employed in municipal or
private conservancy.

§ 53. The Dom and Ghasiya. It was remarked above that in the
Gangetic region there were functions which even the scavenger caste would
not undertake, there being the Dom at hand to perform them. Here, then,
is found a caste which, if not at the bottom of the social scale, is, at
least, not far from it. It is not, however, a scavenging caste by tradition,
nor is it homogeneous. There are Doms and Doms. In the Kumaon and
Garhval Himalaya, the Dom lives by agriculture and village handicrafts.
Further west, the Panjab Dijmna is often, it is true, the village sweeper,
but his ordinary trade is that of cane -work. This last is, in fact, the
occupation most widely spread, on the whole, throughout the caste. The
Dom is at his lowest in the Bengal Delta, whither the caste is said to
have been imjiorted from upper India, to do what no local caste would do.
In Bihar and its neighbourhood to the west, the Dom seem to fall into
two sections. One settled down to village life, mat-weaving, basket-making,
and labour, with a little scavenging thrown in, the other more or less
nomad, and containing gangs said to be expert and artistic burglars and
thieves. Some stray tribes seem to have penetrated across the Central
Belt into the north Telugu country and the Karnatic. In the former they

Castes and Caste-Groups. C. Subsidiary Professional Castes. 85

are coarse weavers, and in the Dekkan, acrobats, dancers and bad cha-
racters generally. Both these bodies have the appearance of belonging
to the Kol-Dravidian race, possibly through the admixture of local blood.
In the same way, the Dom of Dacca, long separated from their native
country up the Ganges, have acquired characteristics different from those
■of the Dum of Bihar. It is now generally believed that the Dom were
settled in force along the southern Himalaya at a very early period, and
judging by the forts and strongholds called after them, they were in a
dominant position, like the Dasyu encountered by the first Vedic immi-
grants. The Dum still on the hills were enslaved by later comers, such as
the Khasya and refugee Rajputs and Brahmans. The community is divided
into four groujis, field-labourers, weavers, and m.etal-workers; cane-workers
and the lower artisans; exorcists, porters and leather-workers, and, finally,
musicians, mendicants, and — tailors. The pom of the plains, when settled,
tend to establish separate castes of cane-workers (BansphOra, Basor), and
labourers. In spite of efforts to get them to work themselves into a better
position they seem to have no aspirations beyond their traditional occu-
pations or a little petty cultivation. But in social intercourse they disown
the nomads. It must be noted that the Diam of the Panjab, whatever their
nominal connection with the Dom, are now an entirely separate community,
both in occupation and social position.

There is a small community called the Ghasiya, which, though pro-
bably not connected with the Dom by origin, may be taken with it in
view of its kindred position and occupation. It has been held, in Bengal,
to be a sub-caste of the Hari, but it appears to be an independent offshoot
of some Kol tribe of the Central Belt, and to have been severed from
its parent stock at a comparatively recent date. The Ghasiya is still divided
into its totemistic exogamous sections, and keeps up the worship of the
field goddesses and other genii of its native haunts. In the neighbourhood
of the larger Kol tribes the Ghasiya occupy but a low position, and
perform on drums and trumpets at festivals with other menial functions.
In the plains, however, the Ghasiya have entered private service as grooms
and elephant-drivers. The caste keeps much to itself, and, low as it is,
it eschews the menial offices imposed upon it in the hills, and especially
avoids the leather-worker and contact with dogs.

C. Subsidiary Professional Castes.

§ 54. This comparatively small group comprises a number of bodies
which, though not so directly concerned with the every-day life of the
masses as those dealt with in the preceding paragraphs, exercise functions
which are intimately connected with certain phases of the domestic or reli-
gious observances of at least the upper and middle classes of the Brahmanic
community in most parts of the country, and stand intermediately, as it
were, between the village and the specially urban castes.

Bards and Genealogists (782,500). These ancient professions are
usually found more or less linked together, and in India the connection
is peculiarly intimate. From the earliest times chants in praise of the
founders and heroes of the clan have been recited to tickle the ear of
the ruling Chief when sitting in formal assembly or heading a procession
through his streets. Still more essential were they in battle, to encourage
the fighting members of the community to emulate or excel the deeds of

86 5 Ethnography.

their ancestors. The annals of such enterprise with the personality' of the
principal performers became, naturally, the special study of those whose
duty it was to set them to verse and directly connect them with the patrons
before whom they have to be recited. The Bard, therefore, developed
into a sort of Herald, and as his office, like all others in India, tended
to become hereditary, the pedigree of those he served was transmitted
in all its ramifications from father to son, with that marvellous accuracy
of memory which is marked feature of the Brahmanic intellect. The im-
portance of such knowledge can hardly be overrated in a country where
the licit and the prohibited degrees of affinity which form the basis of
all arrangements of marriage or adoption, are the subject of most minute
and complicated regulation throughout the community from top to bottom.
In the course of time, therefore, the genealogist more or less split off
from the bard, and took the higher rank at Court. His functions are chiefly
exercised among the Rajjiuts, but in the Panjab some of the Jat clans,
and in Gujarat some of the leading Kanbi families, utilise his services.
As a rule, each of the ruling and leading families keeps its own genea-
logist. The rest of the community is divided into circuits, assigned re-
spectively to a certain member of the fraternity, who annually visits each
family in order to learn what domestic occurrences have taken place since
his previous visit. In modern times every one of these incidents is entered
by him in his register. Such is the reputation of the genealogist for ac-
curacy and knowledge that this register is accepted as final in any question
of affinity or relationship, and even before such "vahi" were customarv,
no Rajput ever thought of disputing the decision of the genealogist upon
these points. The principal caste coming under this head is the Bhat,
sometimes called Bharot in Gujarat and Rajbhat in Bengal. A question has
been raised whether the caste takes its origin from Brahmans who in old
days secularised themselves in order to act as Court poets and panegyrists,
or whether the function devolved upon a member of the Rajput clan to
which the Bhat was attached. There is evidence on both sides. In every
tract in which the Bhat is found, the community contains two sections,
of which the Brahma Bhat is the higher. In Rajputana, the Brahma, or
Birm Bhats are treated as Gaur Brahmans, and in the east ofOudh, that
sub-caste of Brahman which is native to the locality, actually performs
the duties of bard, and sometimes of genealogist. Again, the person of a
Bhat has always been considered inviolable, like that of Brahman. On the other
hand, a Brahman is never known to drop his exogamous subdivision by
Gotra, whilst the Bhat are subdivided according to Rajput custom. The
inviolability of the Bhat, too, may be attributed not only to the character
of herald or privileged messenger or forerunner of Chiefs, but to the
inexpiable guilt of destroying the only recognised authority upon pedigree,
and the apprehension of the vengeance or reprisals that would infallibly
follow such an outrage. It is true that the Bhatrazu of the Telugu country
are subdivided into the Brahmanical gotra, but this branch of the caste
is an exotic, introduced, under the name of Magadha, through Orissa and
probably from Bihar, in the course of invasions of the .\ndhra region from
the north, and has not kept up cither its traditions or its occupation
amongst the once military Dravidian castes to which it was attached. On
the other side, there is the fact that the Bhat is a distinctively Rajput
institution, and, except for the colonies in Telingana and eastern Bengal,
is only found where Rajput influence is supreme. Even in Gujarat, where

Castes and Caste-Groups. C. Subsidiary Professional Castes. 87

the Bhats are numerous, all their sections trace their origin to some part
of Rajputana, and, as a rule, the Bhats in regular employ dress as Rajputs
and have Rajput names. In regard to the distribution of the work of the
caste, the Brahma-Bhat usually takes upon himself the duties of poet and
reciter whilst the others look after the pedigree. In upper India, too, the
former do not take up permanent posts, but are engaged for the occasion.
In Rajputana itself, the male Bhat, it is said, undertakes the care of the
pedigree of the male line, and his wife that of the female. In these days,
the Bhat docs not enjoy by any means the same position as of yore, though
a good reciter has still a high value, and in Gujarat, a popular genealogist
has considerable influence as counsellor in the households of his clients.
Even in the west, however, the Bhat has been obliged to leave his traditional
profession to a great extent for trade and cultivation, like the Bhatrazu
of the south. In eastern Bengal, where the caste is exotic, it ranks much
lower than in upper or western India, though it wears, as elsewhere, the
sacred thread. The Bhat there still practises the profession of genealogist,
and each member of the fraternity has his circuit which he visits annually.
At other times he is in request only in connection with marriage ceremonies,
in which he takes the part of herald between the two houses concerned, and
acts also as go-between in the preliminary stages of the family arrangement.
But in the eastern districts, the Bhat has been reduced even to the trade of
making leaf-umbrellas. Some of the Rajputana Bhat accjuire herds of cattle
and carry salt, grain and piece-goods to localities remote from the railways.

In this respect they fall into line with the Car an, a bard and genea-
logist of a lower type, whose range lies between Kach and Rajputana.
The name seems to connect the caste with grazing, and it is by cattle-
breeding and transport by pack-bullock that the Caran mainly now gets
his living. There is an old and long obsolete connection between the
Caran and the Kumbhar, or potter caste, the link being said to have been
the joint trade of ass-breeding, but the relations have now passed into
the stage of violent but unexplained hostility. It is possible, of course, that
this misty tradition may account to some extent for the inferior position
which the Caran, even when he is exclusively a bard or genealogist, oc-
cupies with reference to the Bhat. The Caran caste is subdivided into
geographical sections with numerous exogamous sub-sections. The families
in permanent employ as genealogists intermarry with each other only, not
as a matter of caste, but, as amongst the Jats of the Panjab, on purely
social considerations. They have thus acquired a physical appearance far
superior to that of the cultivating and cattle-breeding sections of their
community. The profession, however, as among the Bhat, has gone down,
and only a minority now live by it. Most of the western, or Kach, Caran
live by transport on pack-bullocks. Here again their trade has suffered
by the extension of railways acrfiss the desert tracts, but many of them
have adapted their operations to the new order and ply along the main
feeder roads to the chief stations. The Caran who are thus engaged
bear a striking resemblance to the Banjara of upper India and the Dekkan
in appearance, dress and customs. The Banjara of the north have, in
fact, a large subdivision called Caran, and it is possible that there was
of old some tribal connection between them and the Caran of the west,
lost through the migration of the latter.

The Caran shares with the Bhat the reputation of personal inviolability,
and numerous cases are on record, extending even down to 1861, of their

88 5- Ethnography.

killing one of their girls or old women, or inflicting serious, even fatal,
wounds upon their own persons, in order to fix the guilt of certain acts
upon those opposed to them. In earlier times, from at least the is'*" cen-
tury downwards, both castes were the professional securities for the per-
formance of a contract or the repayment of a debt, and no important
document of this sort would be accepted as valid without the "dagger"
and signature of a Bhat or Caran at the foot of it. This practice arose,
ai>i)arently, out of that of obtaining the guarantee or escort of one of
these castes for every caravan or transport train from the coast across
Central India. But the origin of the notiort of the inviolability of the
Caran is as obscure as in the case of the Bhat. The Caran, it is true,
has the reputation of being a violent and turbulent character, whose
ghost is particularly vindictive and malevolent. The curse of a Caran,
therefore, was powerful against one's enemies, and a member of the caste
used to be engaged, like Balaam, to accompany the army of the Chief
to battle, and curse the foe. The women of the caste, too, are reprehen-
sibly familiar with si)ells and charms, and in north Gujarat, the tombs of
some of them are worshipped like those of the local goddesses. On the
whole, however, the sacredness of the office of an authoritative repository
of the family pedigree and achievements seems to be the more probable
source of the conception.

The only other caste which it is necessary to mention under this
head is that of the Dum or Mirasi of the Panjab. The members of this
community are both minstrels and genealogists. Their Brahmanic name
of Dum may have some relation to the former accomplishment, as the
Dom are, as stated in the preceding paragraph, to some extent, musicians.
But the Dum as they exist in the present day are far above the Dom
alike in appearance, position and attainments, though still amongst the
lower classes out of communion with the peasantry and artisans. They
are almost all Muslim, and the name of Mirasi is derived from the .Arabic
for inheritance and may thus be taken to refer to their work as genealogists.
In this capacity they are much below the Bhat, and officiate chiefly in
the families of the lower agricultural population and for the impure castes.
Some Jat families employ them, but the accredited genealogist for that
race, strange to say, is the Sahsi, a criminal vagrant tribe of the province,
whilst the families ambitious of a rise in society engage, as above remarked,
the Jaga Bhat. The musical attainments of the Mirasi" are considerable.
Some only sing, others play the flute, pipe, lute, cymbals and different
sorts of drum. Their women also dance and sing occasionally, but only
for the delectation, it is said, of patrons of their own sex. Those who are
genealogists in permanent employ of a definite circle of clients hold their
office hereditarily, and do not associate or intermarry- with those similarly
engaged among the impure castes. The profession is by no means un-
remunerati\e, especially where agricultural prosperity connotes the neces-
sity of an improved family tree. Even in the open market, the Mirasi is
a popular and well-paid feature of every fair and large wedding. Unfor-
tunately, the Mirasi, like the Bhat in the eastern parts of India, is a
shameless blackmailer, and the refusal or inadequate requital of his demand
is followed by often witty and invariably outspoken burlesques of the
genealogy of the ill-advised recusant. In eastern Bengal, the Bhat, who
there resembles the Mirasi rather than his own namesake of Rajputana,
is said to varv his stock ridicule of the manners and customs of Europeans

Castes and Caste-Groups. C. Subsidiary Professional Castes. 89

with depreciatory references to the ancestry of any local magnate
whose purse-strings may have been drawn tt)0 tightly on the Bard's
last visitation.

§ 55. Astrologers and Exorcists (205,300). The impnrtance of the
horoscc>i)e, or birth-letter, and of a lucky day and hour for each domestic
ceremony is so great in the eyes of the Brahmanic community that the
duty of casting the one and of ascertaining the others is usually entrusted
to none but a Brahman. In many cases he is maintained by the village
for the purpose and remunerated out of the crops, and in most Native
States the JyotisI is an honoured official, endowed with salary and estate
by the Chief His function does not entail any separation from his sub-
caste, so that this class of astrologer does not figure in the census returns.
There is, hcnvever, a much lower grade in the profession, called by the
same name, or rather, by its popular abbreviation, Josi, who is so returned,
chiefly in the upper Gangetic plain and in Central India. He lives by
palmistry, exorcism and omen-reading, and accepts remuneration for aver-
ting the evil influences of eclipses and of the phases of certain maleficent
planets, especially Saturn, and generally pandering to pre-.\ryan credulity.
The subdivisions of the caste indicate, too that the Josi is a community
of very mixed descent, and if connected at all with the Brahman, is only
one of the degraded sections. This seems to be admitted in the case
of the Dakaut, the astrologer of the Jamna valley and Rajputana, who
is of the Agroha stock, unclassed for taking to an unorthodox course
of life. The Ganak, again, of the Brahmaputra valley, are said to have
been cast out by their Bengal fellows for undertaking the duties of family-
priest to the carpenter caste. The Ganak moved int<j Assam, where,
through the influence they acquired as court astrologers to the Koch and
Ahom Chiefs, they settled down into a rank inferior to that of the Brahman
alone. A distinction must be drawn between the Josi of the plains of
upper India and the same caste as found in the Kumaon hills. In the
latter tract the Josi, whatever his position before his migration, has acquired
the status of Brahman in his present home, and intermarries with the
Kanaujiya and other sub-castes. This may be due in part to his worldly
success, as for many generations the Josi has almost monopolised the
sweets of State appointments in Kumaon, and flourished on them. In the
Dravidian country, the profession of exorcist is widely spread, owing to
the prevailing demonolatry, which requires variety of treatment. The
determination of a lucky day, too, probably falls to the priests of the
different communities of the lower classes, and to the Brahman in the
upper. On the Malabar coast, however, there are a few small castes which
appear to be somewhat specialised in these arts. The Kanisan, Panan
and Velan combine exorcism not only with devil-dancing, which is the
usual twin calling, but with herbalism also. Probably all three castes are
descended from the hill tribes of the neighbourhood, but have long been
settled in the lowlands under the protection of the Nayar. In most parts
of India there are specialists in exorcism and protective spells, though
they may not have been yet fofmed into castes. The averter of hail, for
instance, is an institution in parts of Bengal, in the lower Himalaya and
in the north Dekkan. In the Kumaon tracts the duties fall to a special
section of the Dom. In Bengal, there seems to be a caste for the purpose,
called the S'ilari, but it is not returned at the Census. Possibly it has
died out, since it is frankly admitted there that people did not think it

90 5- Ethnography.

worth while to maintain a wizard who could only keep hail off the crops
of his patron without having the power to call it down upon those of
his neighbours. The Garpagari of the Maratha tracts is a distinct caste,
though, like the S'ilari, it is on the wane; not, however, for the same
reason, as the want of confidence now felt in the exorcist is here due
to his inefficiency even as a protector of the crop, without any after-thought
regarding his powers of maleficence. It is worth noting, perhaps, that
these exorcists of the forces of Nature must be remunerated in kind,
never in cash.

§ 56. Temple services, a) Priests (695,400). In treating of the
Brahman, it was mentioned that whilst the post of priest in a family of
a pure caste was one which could be occupied with credit by a member
of the sacerdotal order, ministration in a temple was held to be a duty
only to be undertaken by a degraded, or at least, one of the lower, sub-
divisions of Brahmans. The distinction, it was pointed out, lies probably
in the divergence of the worship of the non-Aryan deities of the existing
pantheon from the old Vedic sacrifices, still held in reverence, at least in
theory, by all orthodox Brahmans. There is also the risk, or perhaps the
certainty, of contamination to be incurred in disposing of the offerings
made in the course of these services. Notwithstanding these drawbacks,
Brahmans are found to perform the necessary offices before the god in
the great majority of the temples of their creed. Equally low in the esti-
mation of the order is the Brahman who subsists upon the fees and of-
ferings of pilgrims at the great centres of religious resort, and still lower,
the Mahabrahman, who takes part in funeral rites. All these, however, are
included under the general title of Brahman. Outside this designation are
some small classes who claim to be Brahmans because they perform temple
service, but who are recruited from the lay castes of the vicinity. The
Pujari and Bhojki, of the Panjab Himalaya are cases of this kind, and,
though repudiated by the Sarasvat Brahmans to whom they have attached
themselves, they seem to have all the position of the order among the
people to whom they minister. The Bhojak and Scvak of west Rajputana,
again, who have been mentioned in connection with the Banya, are held
to be Brahmans, albeit degraded by their connection with the Jain wor-
ship. The real reason for the lowness of their position is surmised to be
their foreign origin, of which mention was made above. The impure
castes, and, in the Dravidian country, a good many of the lower agri-
cultural castes, employ their own caste-fellows for priestly duties outside
the temple, whilst a few castes, in the south, officiate for not only their
own body but for other castes of similar or slightly superior rank. Ge-
nerally, however, these semi-priestly castes are themselves of low rank.
The Panclaram, for instance, is generally considered to be a branch of
the Amli, a fraternity of Tamil religious mendicants; but there is one
subdivision considerably above the average of the latter class, which is
educated to a certain extent, wears the sacred thread, presides over
monastic and temple establishments, and officiates as priests to the great
VeUalan peasantry and the castes immediately below and above it. Some
of the Dasari, too, in the Telugu country, rise far above the rest, and do
service in temples and with respectable families of any caste below the
Brahman. The Valluvan, once the priests of the Pallava dynasties, now
officiate for the PaHan and Paraiyan and have lost much of their former
position by so doing. Like several low castes in various parts of India,

Castes and Caste-Groups. C. Subsidiary Professional Castes. 91

the Valluvan have produced a widely popular poet, Tiruvaljuvan, who is
said to have married into a VeUalan family. It is conjectured that the sa-
cerdotal functions of this caste were superseded by those of the Brahman,
when the latter found his way into the Dravidian region. Now, besides
their cm|)loyment by the castes above mentioned, the VaMuvan have to
look to astrology and herbalistic medicine for their living, and here they
enjoy the custom and confidence of far higher castes. In some villages,
indeed, the Valluvan is on the staff, and receives his annual (juota of
threshed grain from each household. It may be remarked that they do
not ever intermarry with the castes to which they act as priests, unless
they belong to the pure section. The Tambaja, a small caste of temple-
priests in Telingana, hold almost the rank of Brahmans, and where they
have taken to cultivation are still quite in the upper line. It is said that
their name, the local rendering of Tamil, is due to their having been sent
uji from the south by the great reformer, S'ankaracarya, to labour on the
Coromandel coast. As they are mostly worshippers of S'iva, many have
joined, it is said, the Lihgayat community in the inland districts. The true
priests of the latter, however, are the Jarigam, a caste of considerable
influence in the Karnatic. It seems to have been called into being to
satisfy the desire of the converts of Basava to retain priests for their
Dravidian forms of worship after they had split from the Brahmans. In
the tracts where Lingvantism is most powerful the Jaiigam are subdivided
into the usual monastic and secular sections. The former, in turn, are
either stationary in monasteries, or put in charge of a circle of villages,
each of which they visit in turn, imparting doctrine and counsel. In the
outlying parts of the Karnatic, the Jaiigam is not unfrequently a wandering
mendicant of a religious type, living upon doles from every class of the
population. The secular Jarigam, again, is often a trader or money-lender.
The Census returns of this caste, though possibly fairly accurate in the
aggregate, are defective in detail. In the south Dravidian districts, that
is, the term Jarigam is used of any Liiigayat, whilst in the north on the
contrary, many Jarigam are returned as Lirigayat or as Virsaiv Brahman.
A small caste corresponding somewhat to the Valluvan, is found in Gujarat
and the north Dekkan, called the Garuda, which serves the leather-
working castes as priest. In some parts they eat with their clients, but
in Gujarat they are generally superior to the latter in education and
physical appearance. From one of their subdivisions it might be surmised
that they are the descendants of a superior class driven out of Rajputana,
like so many others in the west. The Garudi of the Maratha country is
of a lower type altogether, and belongs to the Maiig caste.

In the Panjab, there is one caste requiring notice, the Bharai, which,
however, is often returned simply as Sekh. The Bharai is the special
guardian and ministrant of the shrine of the popular Saint Sakhi Sarvar,
of the Indus. Whether he is, since his canonisation, Muslim or Brahmanist,
it is difficult to judge from the crowds that throng to his tomb; but the
Bharai are of the former creed. They haunt the centre and submontane
parts of the Province, and live by conducting pilgrims down to the shrine
at Nigaha, in the Derajat. It is said that some of the Bharai have taken
to music and call themselves Mirasi. The only other occupation with
which they are associated is circumcision, which rite they perform in
supersession of the barber on the lower Indus. Along the I'athan frontier,
there is a body, incoherent and multifarious, which locally arrogates to

92 5- Ethnography.

itself the title of Ulama, or the learned. The entrance-qualification,
however, appears to be only the knowledge by rote of a sufficient number
of texts of the Kuran to serve as spells or curses for the practical pur-
poses of life. On the other hand, the term may include the highly educated
Maulvl of the city mosque, and the Kazi, who may or may not be erudite
in the law he administers. It is not, however, a caste, and as a functional
body, enjoys as low a reputation for piety as for erudition, and is the
subject of many biting proverbs along the frontier.

§ 51. b) Temple-servants (300,500). There are certain castes in
almost every part of India, but especially in"the south, which are dedicated
to offices within the temple other than those of actual worship. They
wash the images of the god, deck it with flowers, and keep the precincts
clean. Most of them have other and more secular avocations, generally
connected with leaves or flowers, such as umbrella-making, the preparation
of leaf-i)latters for Rrahmanic festivals and garlands for ceremonial use.
The caste most widely spread of all thus engaged is the Mali, or garland-
maker; but as nearly the whole of the caste is in the present day occupied
in gardening or agriculture, it has been reviewed already under the head
of special cultivation. In Bengal there is still enough of the traditional
work left to justify a separate subdivision to perform it. In other provinces,
too, the growth of flowers and the making of garlands, particularly those
for the temple, are the work of special bodies, but they are generally
distinct from the Mali. Such arc the Phul-Mali, Phularl, Hugar, and the
like. It is still necessary to be specially brought up to the trade, lest
mistakes be made which would be ruinous. One god has to be decked
with flowers which are abhorrent to another; certain flowers, too, are
required by convention for certain occasions, and the marriage-coronet
must contain the prescribed flowers and no others. The small castes
above mentioned are generally found south of the Vindhya, in connection
with the caste of Gurao, which is accredited to certain temples, usually
those of S'iva, where the post is permanent and hereditary. The Gurao
also make the leaf platters required for caste-feasts and other banquets
on a large scale, a task which in upper India is performed by the Bari,
who, however, does not serve temples. In the Telugu country, the Satani
does the work of the Gurao and a good deal more, for it appears that this
caste was brought into being to aid the propaganda of Ramanuja, its patron.
It is associated, therefore, more closely with religion than a mere temple
servant, and acts as priest to several other castes in a good position,
as well as the lower classes. In contradistinction to the Gurao, the Satani
is Vaisnava, and those of the caste who are brought up as priests are
fairly conversant with the Puranic authorities of their sect. Formerly, the
Satani called in Brahmans for their ceremonies, but of late their own
priests have come into favour. The Balija community generally employ
the Satani, but those who are redundant in this capacity, take to umbrella
and garland making. The Tulu caste of Dcvadiga is not found outside
Kanara, and where not engaged in temple service, the caste has taken
to cultivation and the lower grades of State service. The curious trans-
formation of the Barber into the temjjle servant in Malabar has been
already mentioned, and there are about 8,000 of the Marayan who combine
that duty with the manipulation of the tem])1e drums when required.

§ 58. Dancers, Singers &c. (135,900). That these professions should
be placed immediately after those connected with temple-service is by no

Castes and Caste-Groups. C. Subsidiary Professional Castes. 93

means so anomalous as regards sequence as it may appear at first sight.
In India, as in other oriental countries, dancing and singing are profes-
sional accomplishments or ceremonial observances, and only among some
of the wilder tribes is the dance a form of private recreation. In Brah-
manic circles there are recognised dances, generally of a religious signi-
ficance, danced among women, and, also, a few highly heterodox tripu-
diations associated with the rites of some particular sect, and ignored by
the rest of the community. There are, again, the sword-dances of the
Khattak and other frontier tribes, and most of the more primitive forest
communities have their reel or square-dance with its traditional figures
handed down as a tribal possession. With these exceptions, the dance
in India is a performance by trained professionals, of a character which
may be called posture-singing, or illustrating by gestures the words sung
by the performer. The subject of the ode, except amongst the Muslim,
is usually connected \vith religion or mythology. In the Dravidian region
the dancing takes place within or before the temple, in honour of the
god, especially of S'iva in his many forms, and the performers are de-
dicated to him and form part of the establishment of the temple. The
women have their off-time, of course, which is spent in the practice of
the ancient mystery everywhere, now as of yore, associated with profes-
sional dancing. In fact, the old Dutch travellers when introduced to these
bevies, did not mince their words, but habitually refer to them under
the title of "danshoer", an appellation even more applicable to the dancing
castes of the north than to those in the south, since the former have no
connection with religion beyond the dedication of the individual to the
worship of a certain god, if she be of the Brahmanic faith. It is worth
noting that owing possibly to this connection with the popular pantheon
or, as some think, to the more distant tradition of communistic marriage,
the courtesan is not a degraded member of Indian society, but, like the
Hetaira of Athens, is not only tolerated, but respected. There is, of
course, every grade to be found amongst them, as in all countries where
inequality of purse is the rule, from the ragged nomad in her filthy little
reed-booth, whose musical and terpsichorean attainments are of the lowest,
to the highly-trained singer of the great city, well versed in alike clas-
sical and popular poetry, w'hose diction is often quoted as the standard
of Hindi or Urdu polite conversation. It may have been gathered from
what has been stated above, that the two arts, dancing and singing, go
together, and comparatively few and those only of the highest rank, sing
without the plastic accompaniment. Recruited as they are from all castes,
Brahmanic and Muslim, under a number of titles, honorific or the reverse,
it is not worth while to dwell upon them here otherwise than cursorily,
as an ancient and recognised grade in Indian society. In upper India,
alongside the functional titles of Tavaif, Kasbi, Xaik, and so on, there
are found the semi-religious designations of Ramjani, Gandharp, Ras-
dhari and the lower ones of Kancan, Besya, Patiiriya and so on. In
the west, besides some of those above mentioned, there are the Naikin,
of the semi-religious type, and the Kalavant. The religious establish-
ments referred to above are all in the south. In the Dekkan there are
several of comparatively small renown and endowment, and only nine
women have returned themselves under the specific name there given to
them. Even in the Tamil country, where the accommodation for this class
round the chief temples indicates the extent of the community in old

94 $• Ethnography.

times, the number returned is far below the actual, since many of the
girls give the name of the caste in w*hich they were born, instead of
that to which they were dedicated when they wedded the gr)d. It is the
duty of the Dasi to fan the god, present to him the sacred light, and to
sing and dance before him when he is carried in procession. Owing to
their Hrahmanic connection, they do not consort with the Kammalan, or
artisans, who belong to the Left hand, nor, of course, with the impure
castes. Their sons become musicians, often of considerable skill and learning,
and occasionally marry into respectable castes. The daughters folhnv their
mother. In the Telugu country, the caste is known as Bogam or Sani,
and is widely scattered in small numbers. There is only one institution
of the sort common in the Tamil region. The Kanarese Devali are mostly
ascribed to a god or to temples, as in the south. B(nh here and in Te-
lingana, the recruits arc from the Palli, and Iloleya, but on the coast, the
breed is apparently from a fairer stock, like the Tiyan, or bastards of
the Ilavik. All these dancing and singing castes have their strict rules
about initiation, conduct, inheritance, and the observance of caste re-
gulations, enforced through a caste Council, or Pancayat, like the larger
communities.

D. Urban Castes.

§ 59. The majority of the castes coming under this head are here
placed not on account of any ethnic distinction between them and those
already described, but merely in consideration of the generally urban
character of their occupations. Most of them, indeed, are but offshoots
of larger bodies still unaffected by the influences of the city, and are
finding their way back to the village as communication grows easier and
the convenience they represent gets to be the object of a more effective
demand. It should be understood, therefore, that these castes are not
entirely confined to the towns, though it is there that they find at present
the main field for their labours. They may be conveniently grouped as
shopkeepers, artisans and domestic servants.

§ 60. Grocers &c. (825,000). Under this head come the retailers not
only of spices and condiments but of perfumery also, the functional name
of the Ranya who sells the former in one part of the country being the
same as that of the extractor and seller of scents in another. The latter
is but scantily represented in the Census returns, and is usually a Muslim.
The large proportion of Brahmanists coming under this title may be taken
to be grocers returning their professional, in place of their caste, name.
The Gandhi or Gandhabanik of Bengal is generally a druggist as well as
the vendor of condiments, and when he sells sandal-wood and other fragrant
articles which enter largely into domestic worship, he rises in position.
The whole caste, indeed, pays homage in the spring to Gandhesvari, the
goddess cf perfume, a manifestation of Durga. The Gandhabanik also sells
drugs, and is reputed to be well acquainted with all local medicinal pro-
ducts. A few take out licenses for the sale of opium and intoxicating
preparations of hemp, but the actual sale of such articles is left to a
Muslim assistant. The grocer of the upper Gangetic region generally be-
longs to the Kasar- or Kesar-vani orKasaundhan castes, both some-
what low branches of the great Banya order. The latter derives its name
from dealings in brass or bell-metal, and the former probably from saffron.
Both now sell grain, salt and other commodities which their Bengal con-

Castes and Caste-Groups. D. Urban Castes.

95

freres avoid. Both employ the same caste of Brahman and follow to a
great extent the teachings of Ramananda, and in Bihar, the Nanakpanthi
doctrines. In the Dekkan and west, the Gandhi is not a separate caste,
but merely a petty trader^ of the Vania caste. Subsidiary to this group
may be mentioned the Kunjra, or green-grocer of the north. It is not
a caste, i)roperly so called, except, perhaps, in Oudh and along the upper
Ganges, where the sellers of vegetables are all Muslim and have banded
themselves into an apparently endogamous community. It was stated in
connection with the growers of the bitel-vine that the im])<)rtance of the
"bid" or "birii" in society was held to entitle those connected with it to
a quite respectable position, above that indicated by the ancestry or wealth
of the castes in question. The Tamboli is the caste which sells the leaf
in almost every province except in the south. In Bengal and Bihar, the
caste is supposed to be connected with the Banya and in the Dekkan with
the Kunbi, but in upper India it appears to be a branch of the Barai or
grower of the vine, and in some places the latter sells the leaf he grows.
This, however, is exceptional, but the names of the subdivisions of the
two castes indicate former relationship if not identity. Occasionally the
Tambnll extends his dealings to snuff and tobacco, and even to grain and
lime. In Bengal some of them hold land, but cultivate through hired labour.
In those tracts it ranks lower than up the river, where it sticks to the
shop, and is considered equal to the middle-class peasant in position.

i;6i. Grain-parchers and Confectioners (1,645,200). Both these are
important functionaries in town life in Bengal and upper India, but are
in comparatively little request south of the Vindhya, where the diet and
rules connected therewith are different. The origin of these castes is not
clear, except in the Panjab, where both the Bharbhiinja, or grain-
parcher, and the Bhathiara, or public cook, are of the Jhinvar, or water-
bearing caste. The Bhathiara is only found in the Muslim tracts, except in
the larger cities, since the Brahmanic rules of living do not admit of the
common oven. The grain-parcher is of more mixed origin. In the north,
one of the sub-castes is connected with the Kayasth, and the same re-
lationship appears in the communities of Bihar and the Dekkan. On the
other hand, the Bharbhiinja is often held to be only an elevated branch
of the Kahar, a view that coincides with the known facts further west,
and is corroborated by the existence of sub-castes connecting the com-
munity with the Gohr'.ii, a fishing caste of quite a different part of the
country, and with the Kandu, the sweetmeat maker. In Bihar, in fact,
the Bharbhunja, is considered to be a sub-caste of the Kandu. Towards
Agra, however, the latter takes a higher place, and is almost equal
to the Banya, exclusively engaged in the traditional pursuit of con-
fectionery; but of its numerous sub-castes, some, like the Gorirhi, work
in stone, and others parch grain, like the Bharbhiinja. It seems probable,
therefore, that both the castes originated amongst the fishing and porter
community, and have been reinforced by occupational subdivisions formed
locally to meet a demand for their services. The Halvai, another caste
of confectioners, is entirely distinct, and, in upper India, is often Muslim.
It is a composite body with a good many endogamous sub-castes. One of
these shares the name of the Godiya, or Gufia, the confectioner caste
of Orissa, though without any other connection. In Bengal, the Mayara
caste is like the Kandu of the north, recruited from various bodies and
is subdivided, accordingly, into both Brahmanic gotra and totemistic exo-

g6 5. Ethnography.

gamous sections. Some of those castes have betaken themselves to hus-
bandry, but in that capacity, curiously enough, they will have nothing to d.j
with the cultivation of the sugarcane or the preparation of molasses, the
stock-in-trade of the rest of the caste. In connection with this group of
castes it may be remarked that the upper and middle classes of Brahmanical
society, wherever the caste-system is strictly maintained on the northern
Indian model, are prohibited from eating anything but parched grain or
sweetmeats when on a journey away from their domestic cooking-place;
and this rule may have a good deal to do with the consideration which
is allowed to communities of such mixed Or dubious origin as those which
purvey these convenient provisions.

Sj 62. Butchers 1701,800). No such credit, however, is attached to
the sale of meat, which, naturally, is chiefly in the hands of a flesh-eating
community like the Muslim. It is not to be supposed, from this that
Brahmanists are universally either vegetarians or fish -eaters. Customs
differ in this respect in different parts of the country and amongst different
castes. Beef and pork, indeed, are eaten by none but the lowest of the
community, but in the middle classes, especially in the Dravidian country,
the consumption of mutton and goat is considerable, though the mediation
of a professional salesman, except in the towns, is comparatively rare.
In Vedic times, the Arya were apparently accustomed to eat meat,
and acquired the vegetarian habit as they got acclimatised to the
tropics. Nowadays, the only butcher caste not Muslim is the Khatik,
and this community, though breeding pigs in the north, only slaughters
sheep and goats, the skins of which are tanned by its household. In
the south, the Khatik is merely the professional title of the Muslim
mutton butcher. The Kasaf, or Qasab, of upper India is almost exclusively
Muslim, and in the Panjab is merely a functional branch of the TcU, or
oilman. Further east there are subdivisions, and that which deals in mutton
holds itself above the beef-butcher. This last is, of course, anathema to
the Brahmanic world, and in some places is "boycotted" by tradesmen,
so that it is obliged to make its purchases through the intermediary of
one of the lower Brahmanic castes.

§63. Pedlars and Glassworkers (424,100). There is a certain con-
nection between these two apparently incongruous occupations. In the
north, especially, there are several small castes which go round with beads,
glasswork, bangles, and so on, which, if not made by themselves, come from
the hand of those nearly related to them. Others deal in haberdashery,
small hardware, soap and mirrors. Even if they were not castes at the
outset, they all tend to become such, with subsections and regulations as
to marriage and the like, independent of the communities to which they
originally belonged. The Bisati, a Muslim body, is an example of this
tendency. The Ramaiya, or Bhatra, of the east Panjab, however, seems
to be a tru2 caste, hailing from MarvacI or the neighbourhood, and having
conceded to it the rank of a low Brahman. It is allowed to wear the thread
and to take offerings at eclipses. Otherwise, the Ramaiya tell fortunes and
invoke upon almsgivers blessings which have the reputation of being ef-
fective. They are by tradition petty traders, and in that capacity travel
far and wide, even south of the Vindhya. The caste is much scattered
and is only found in strength in the Bijnor district of Rohilkhand, far
from its original home. But the Ramaiya always regard themselves as na-
tives of the Panjab, and most of them are Sikhs by religion, though

Castes and Caste-groups. D. Urban Castes. 97

employing Brahmans as priests and Brahmanist barbers in daily life, thus
showing a considerable laxity in their faith. The Manihar is strictly the
maker of spangles for the adornment of glass bangles, but in some places,
as in the Panjab, the caste make the bangles themselves. The Curihar,
who follows the same trade, is a separate community, but both travel
about with their goods and do not keep shop. Both, too, are for the most
part Muslim. The Kan car, who also works in glass, takes the place of
these castes in the Dekkan and west, and the Lakhera, a northern
caste, makes the same sort of ornaments in lac. In the Dravidian region,
the corresponding caste is the Gazula, a sub-caste of the Balija, of low
position. In the Tamil country- it is called Vajaiyal, and is taken as a sub-
division of the Kavarai, the Balija colony of those parts. The Cudigar,
generally a Muslim, is probably the Curihar of the north settled in the
Dekkan. On the Orissa coast is a caste called Patra, or Pator, which
peddles silken necklets and cords, like the Patva in other parts. Finally,
under this head may be included the makers of conch armlets, who are
a caste only in Bengal, where they are called S'ankhari. It appears that
through the Subarnabanik they have some connection with the gold-
smith castes; but they deal exclusively in the armlets made from the conchs
brought from the Gulf of Manar. Similar armlets are used in other parts
of India, but they to not seem to be made by a special caste.

§ 64. Artisans, a) Tailors (867,800). Throughout upper India the
tailor's craft is exercised by a composite body, nearly half of which is
Muslim, recruited, judging from the titles of the subdivisions, from many
sources, not all of the lowest. In the Panjab the Darji is merely a
functional name, and in each large city the tailoring body is governed
by a craft-guild. If any caste can be said to produce the tailor more than
another it is the DhObi or washerman. In the Gangetic region the Darji
regulates his life on the model of the upper Brahmanic castes, and one
of the principal sub-castes bears the name of Kayasth. But the caste is
not popular, any more than it used to be in Europe, and is the subject
of similar depreciatory proverbs. Its work is badly paid, but the Darjf
rarely looks out for more lucrative employment. The general style of
dress amongst the peasantry in the greater part of India renders the craft
unnecessary, so the caste is mostly congregated in the cities. It is sub-
divided according to the general nature of the work undertaken, and is
then split up into more minute sections. The repairer and darner is at
the bottom, and amongst the Muslim, tent-making stands high, as being
the occupation of Ibrahim (Abraham!, the patron of the craft. Turban-
making, too, is honourable. In the west, indeed, where the latter article
of attire is more elaborate than in the north, and each caste has its own
distinctive form of head-gear, the turban-folder is a separate community,
and ranks high amongst the Darji. In the Dekkan the S'impi is often a
travelling piece-goods dealer, going from village to village with his pack
upon his pony. He also traffics in small pecuniary- advances, and this is
perhaps the reason for his figuring in bad company in the village rhymes.
One of the popular religious teachers of India, Namdev, belonged to this
caste, and several of the sections of the Darji and similar castes are named
after him. It seems as if the Dekkan tailor were more allied to the
lower trading classes than to the rest of his craftsmen, and certainly he
follows the traditional employment less than any of them. The Gujarat
Darji, too, seems to have sprung from one of the lower classes of traders

Indo-.\ryan Rciearch. II. 5. ,

qS S. Ethnography. |

of west Rajputana, to which locality he claims to belong. Like the S'impi,
he lives after the manner of the upper middle classes, and is strict in
his religious observances, though alleged to be addicted, like the gold-
smith, to helping himself too freely to some of the material entrusted
to him to make up. In the Dravidian districts there is no special caste
of this sort, the tailors in the cities being all Muslim. The introduction
of sewing-machines, and the growth of the fashion of wearing cut-out
garments have tended to the advantage of the town Darji, and even in
villages the machine is often to be seen enstalled amid surroundings of
apparently the most incongruous simplicity.

§ 65. b) Dyers and Calenderers (495,0001. The calico-printers, calen-
derers and dyers appear to be connected remotely with the Darji castes,
except in the Panjab, where the Chipa is an offshoot of the Dhoba or
washerman, who occasionally does the work of dyeing in madder, though
he leaves indigo to the Muslim Rangrej. Elsewhere, the Chipa stands higher,
and in upper India claims to be descended from some Rajput or kindred
tribe in Malva. The Bhausar of Gujarat, too, admits his connection with
Rajputana; but, though not disowning the Chipa of Agra, asserts his
origin to have been through a Vania caste of the west, and will not
acknowledge relationship with the Chipf of his present province. The
Bhausar, like the Vania, has a Jain as well as a Mahes'ri, or Brahmanic,
sub-caste, and lives much on a par with the trading classes. In the Ganges
valley a good many of the Chipa are followers of Namdcv, the Dekkan
S'impi, a fact which indicates something more than merely sectarian
sympathies, considering the restricted social field of the acceptance ol
these doctrines. The Rangrej, Rangari, or Nilari, workers in indigo,
are chiefly Muslim in the north. In the Panjab this is due, as above in-
dicated, to the abhorrence of the Brahmanist of those parts for the unlucky
colour, blue. In Bihar there is not this prejudice, and the Lilua works in
the local material. In the Maratha country, too, the women wear blue in
preference to any other colour, but here, again, the dyer is usually a Muslim.
In Gujarat, the taste is in favour of more varied colours, and the Bhausar
works impartially in all, except indigo, which is the monopoly of the
Galiara sub-caste. The Muslim engaged in the occupation began, no doubt,
as a functional body, but are now, it is said, closing their caste to outsiders,
and keeping to their own sectional divisions. In the Dravidian country there
seem to be no special dyeing castes, the work being done in the Telugu
country by Maratha Rangari. Plain white with a simple coloured border is the
usual colour worn by the women in both theTamil districts and in lowcrBengal.

§ 66. c) Cotton-scutchers 1760,600). Those who follow the occupation
of cleaning cotton arc mostly Muslim, under the functional title of Penja,
Pinjari, Dhuniya, Behna, or even the Persian, Nadaf. They are mostly
converts from Brahmanic castes like the Tcli or oil-pressers, and those
who have remained in their former creed follow the teachings of Namdcv,
the S'impi, like the Tailors and Dyers, and in the Panjab, the Dhoba.
In the north, where the calling has become the work of a caste, those
who do not engage in it keep shops for the sale of haberdashery, spangles,
bangles, caste-marks and so on. The Muslim, as in similar cases of other
castes, have not altogether abandoned their Brahmanic customs or worship,
and follow the traditions of their neighbourhood in this respect.

§ 67. d) Distillers and spirits-sellers (1,725,000). The traditional
connection of these castes with the provision of a forbidden article, places

Castes and Caste-Groups. D. Urban Castes. 99

them very low in society, in fact, little above the oilman. On the other
hand, since the regulation of the liquor trade has been undertaken by
the State the restriction upon sales has thrown a good many of the
caste on to other occupations in which they have prospered far more
than if they had kept to distillation. In the whi)lesale trade in piece-goods,
timber, salt, etc., the Bengal Suriri is said to have reached ([uite the top
of the tree, and being ambitious of a commensurate rank in society, is
forming a separate caste calling itself S'aha, or Saha, in order to sever
itself as far as possible from the branch which still deals in liquor and
serves in the State distilleries, or takes licenses for the sale of intoxicants.
Others of the caste engage in the boating trade, but will only ply on craft
which are manned exclusively by their own comrades. In sjiite of the
rise in their worldly circumstances, the Sunri have been unable to conquer
the prejudice against them, and have to maintain barbers and washermen
of their own, since the Nai and Dhobi decline to serve them. Even the
Bhuinmalf, who will sweep for them, refuses to accept food from their
hand. In upper India there is the same subdivision of the Kalal caste;
those who have taken to trade severing themselves from those who stick
to the traditional calling. But the Kalal in Bengal will make, but not
sell, liquor, whereas in the north the caste does both. In all probability,
in Bengal the castes are both composite, created as the need for their
services became pressing, whilst in the Panjab and its neighbourhood the
caste is older and more homogeneous. The Sikh connection of the Kalal
or Kalvar, in the Panjab, gave the caste a great lift, and one of the
most powerful leaders of that faith, before the rise of Ranjit Singh, belonged
to the Kalals of Ahlii, and laid the foundations of the well-known State
of Kapurthala. Hence a good many of the Kalvar of the province use
the title of Ahluvalia for their caste. On the other hand, in the west
and central Panjab they have preferred to throw in their lot with the
Pathan, and have elongated their name into Kakkezai. The trading branch
in those parts deals in boots and shoes, bread and vegetables, articles
which the ordinary Khatri considers beneath him. In the south, the Kalal
is found in comparatively small numbers as a distiller, but here he has
to compete with the local Parsi in both making and selling spirits.

§ 68. e) Domestic servants (698,800). The majority of the castes
which traditionally engage in service about the houses of those above them
belong, as already stated, to the fishing and porter communities, whose
touch does not contaminate. The households of the Christian or Muslim,
again, are on a different plane, and must be served by Muslim or members
of the impure castes. The water-bearers, too, who ply in the streets or
from house to house, irrespective of caste, are usually converts to Islam,
or of the fisher caste. If the former, they are known generally as Bihisti,
and form a caste of their own, with functional subdivisions, according to
the water-bag they use or the beast of burden they employ. In some
parts of India, again, there is a caste which lives by rice-pounding for
large families, a work which elsewhere is done by the women of the family.
The small community of Kuta, in Rohilkhand, and of Gola, in Gujarat,
are examples of these, but both are probably branches of some larger body,
the Kuta, perhaps, of the Banjara, and the Gola certainly of a Rajputana
caste. The castes which distinctively belong to the group under considera-
tion, however, are those which have grown up under the protection of
the households they serve, and in most cases are in practice inseparable

lOo 5- Ethnography.

from them. The Rajput families, for instance, used to receive the daughters
of lower castes around them, bring them up in domestic servitude, and
practically own the offspring resulting from the relationship. The link was
in some cases closer than in others, and the males were allowed to marry
outside the household, especially in the Dravidian region. But the bastards
usually became a caste by themselves, living on the bounty of their pro-
tector and employed in duties about his estate or Court. The Gold and
Cakar of Rajputana are of this class, though, as just remarked, some of
the former have moved south and set up for themselves in Gujarat as
rice-pounders. The Khavas of the western peninsula are of the same
origin and position as the Gola, but rank considerably above the latter,
and arc employed in posts of confidence which give them much influence
in the neighbourhood. The girls serve the Rajputni, and some of them
are generally included as part of the dowry when their young mistress
is married off. In Orissa, the Khandait keep Casa girls, and the offspring
ranks according to the caste of the father, as Khandait, Kayasth, etc., the
whole body being known as Sagirdpe.sa, with endogamous sub-castes de-
termined as above. In Bihar, too, there are corresponding communities
which are gradually forming themselves into separate castes. In Eastern
Bengal there is a larger caste of this sort, known by the non-committal
title ofS'udra or S'udir, or, in some parts of the province as Ghulam or
Bhandarf. They are descended from comparatively low castes which sold
themselves to the Kayasth, a relationship which, tacitly though illicitly
still subsists. The caste is nominally endogamous, though amongst families
which are still attached to Kayasth households intermarriage with members
of the latter caste is not uncommon, but the title of S'udra is dropped in
the next generation in favour of that of Kayasth. In the south, the Telugu
Velama and landlords of other castes have a similar institution, the results
of which are known as Khasa, or private property, and are crystallising
into a caste. In the south Tamil country, the Tottiyan have families on
their estates which are already a caste, known as the Parivaram, the
members of which cannot marry without the consent of their lord. In this
case, however, recruits are taken from Paraiyan and other low castes. The
Kotari of Kanara, also domestic servants in local families, are apparently
of the Banta caste originally, though now severed owing to their connection
with the landed interest. It must be remembered in connection with all
these domestic classes that the status of slavery in which they originally
dwelt no longer exists; nevertheless, as has been remarked above with
regard to the predial serfs, the tie between them and the family they
serve retains a great deal of its former character, and is perpetuated
voluntarily by both personal attachment to the household and the benefits
derived from the protection afforded, and also the general tendency of Indian
communities to look upon what has once been as pre-ordained and here-
ditary. The position they hold is recognised and established, and in their
eyes there is nothing to be gained by abandoning it for another, indepen-
dent but precarious.

E. Nomadic Castes.

§ 69. Carriers (897,800). The two great divisions into which this

group naturally falls are those of the pastoral tribes and the Gipsies. The

bulk of the former have been already mentioned in connection with the

function of providing the vast number of cattle required by the village

Castes and Caste-Groups. E. Nomadic Castes. loi

community for the plough and for milking. These, for the most part, are
either stationary, or, when they move, merely camp for a few months
of the dry season on recognised grazing grounds not far from their village.
The Caran, mentioned in connection with the duties of Hard and Genea-
logist, is, undoubtedly a nomad in some of its sections, and, in this respect,
it shares the habits of the Banjara, to whom it is probably akin. The
latter, with its branch known as the Lavana, Lambhani, Lambadi orLabana,
is the great bullock-dealer and carrier by pack-animal for the whole of
upper India, and colonies of it have settled in the Dekkan and as far
south as Mysore. The use of bullocks as a means of transport is an ancient
custom in India but it received its great impetus from the Muslim invaders,
who engaged large gangs of Banjara to accompany their forces from north
to south. Similarly, the British armies in their earlier campaigns trusted
to the Banjara trains for their commissariat and forage supplies, and found
the Naik, or gang-leaders, fully up to the work and worthy of confidence,
it is not certain how the Banjara came to be settled in Rohilkhand and
its neighbouring Tarai, but their own tradition is that they belong to north-
west Rajputana, and were driven out of their native country. They also
once settled in Oudh, but were displaced by Rajputs. In their present ca-
pacity, however, they emerged into notice from their Tarai home. The titles
of their subdivisions, which are very numerous, indicate in some instances,
a desert origin, a hypothesis which is borne out by their appearance.
They are usually a tall, sinewy race, their women especially being re-
markable for their powerful physique. Their dress, too, is that of the
west rather than of Hindustan, and one of their sub-castes bears the
distinctively western appellation of Caran. The Lavana, again, another
section, indicates by its connection with salt a trade from the coast or
Sambhar lake. The colonies above referred to appear to have been left in
the south after expeditions by various Muslim leaders across the continent
to the Dekkan and Karnatic. The settlers seem to have made no attempt
to regain the north, but acquired land, and to some extent adopted the
vernaculars of their neighbours. It is said, however, that the primitive
customs and beliefs of the tribe are more carefully maintained by the
Dekkan than by the North-country Banjara. Other branches are found in
Central India and the Panjab. One section has been converted to Islam,
under the name of Turkiya, a title which has led, by one of the humours
of the Census, to its being numbered amongst the Osmanli and other Turks,
though the farthest region to which it ascribes its origin is Multan. In
the Panjab, too, a good many Banjara are called Sikhs, but this refers
to the creed of Nanak, rather than to the more exclusive doctrines of
Guru Govind. Nanak, indeed, is one of the names most revered amongst
the Banjara even as far as the Dekkan. In upper India some of the tribe
have settled down to trade and money-lending. The Vanjari of the Ma-
ratha country-, too, are to a great extent cultivators, and for some gene-
rations have been scarcely distinguishable from their Kunbf neighbours.
The traditional calling of the tribe has been greatly curtailed by the
extension of railway communication, but a good business is still done,
especially where it can be combined with the rearing and sale of stock
to the peasantry', as in Oudh and upper India generally. In the tracts
where the gangs are organised for travel, the old system of Tahda, or
gang-circuits is retained, and no Tahda is allowed to journey over the
sphere allotted to another. In the Dekkan, indeed, the partition is said

I02 5- Ethnography.

to be not unconnected with predatory excursions by the lower class of
Banjara. The Lavana, under its various designations, is sometimes treated
as a separate caste, and is not often found alongside of the Banjara.
But it appears to be nothing more than one of the older divisions of the
main community, which has kept to the west and south. In the Karnatic,
for instance, the title Banjara is unknown, and the Lambadi, or Lambhani,
occasionally called Sukali, pursues its avocation alone, though on a lower
plane than his comrade in the north. He maintains, however, his reputation
as a cattle-doctor, as well as that of an expert in sorcery and witchcraft.
This last attribute is acquired, it is said,' in the course of a wandering
life, exposed to all weathers in jungles and other unhealthy localities.
Strange diseases make their appearance only to be accounted for by the
agency of witchcraft, and the old women of the Tanda, accordingly, go
in considerable risk of their lives. In compensation, perhaps, the Banjara
is the only caste in which the women are said habitually to take the big
walking-staff to their husbands. There is a small caste, the ThOri, which
performs in the lower Himalaya the duties of carriage undertaken in the
plains by the Banjara or Lavana. They are connected with the latter, and
apparently ply their trade in the same tracts in north Rajputana, of which
tract they say they are natives. But there is another caste of the same
name which is allied to the Aheri, if not identical with them, and these
are altogether lower in rank and pursuits, being mostly fowlers, or at best,
mat-makers, along the Indus. In Central India and the north Dekkan, even
as far as Mysore, there are still a few bands of the once noted Pendhari
freebooters, now engaged like Banjara in the carrying trade. Originally,
the Pendhari were no more than a collection of all sorts of foreign Muslim
disbanded from the Delhi army, and linked together for the common pur-
pose of raiding villages and travellers. They are now a small caste by
themselves, and give little or no trouble to the police. They have a Chief
who rules a small State in Malva, but there is no longer any bond bet-
ween him and the wandering gangs.

§ 70. Shepherds and Woolworkers (4,265,600). These two occu-
pations go together, and are exercised by several communities of con-
siderable numerical importance. Their social rank varies a good deal, but,
in spite of alleged descent from the Jadav family of Mathura which some
of them claim, they stand, on the whole, lower than the breeders of horned
cattle. There are, however, exceptions, such as the Gadi.li of the Panjab
Himalaya, who are of the same stock as the Khatri, and rank but little
below the Hill Rajputs. They are admittedly wellborn, and state that they
were driven from northern Rajputana by the Muslim and took refuge in
the Kangra and Chamba hills. Like all the shepherd classes, they weave
the wool of their herds, both sheep and goats, into strong homespun and
blankets. They are also credited with being very skilful and industrious
cultivators ..f the upland regions affected by them. They have no connec-
tion with the caste of the same name along the Jamna, which is Muslim
and a branch of the Ghosi, mentioned above amongst the cattle-breeding
castes. The chief shepherd caste of the Ganges valley is the Gadariya,
or Gareri, as it is called in Bihar. In that Province it ranks higher than
in the west, but its home is alleged to be in the latter, and some of its
divisions derive their origin from Maratha shepherd clans who came north
through Malva and Gvalior. One of the chief shrines at which the caste
worships is in the last named State. The name of the caste is said to come

Castes and Caste-Groups. E. Nomadic Castes. 103

from the Sanskrit name of the country, Gandhara (or Kandahar) from
which the animal was said to have been introduced into India. Judging
from physical appearance, however, the Gadariya and, except the Gaddi,
the shepherd castes generally, have much more Kol or Uravidian blood
in them than the western cattle-breeder. The Dhangar of the Maratha
country, indeed, is by some identified with the Dhangar, or Oraon, of the
eastern portion of the Central Belt, a large tribe which is thought, mainly
on linguistic grounds, to have pushed its way up north from the Karnatic.
Even in the present day, too, the Dhangar build their shrines in the same
way and of the same sort of unhewn stones as the Kurubar of the Kar-
natic, a once dominant tribe of the south, to which belonged the Kadamba
dynasties of Banavasi in Kanara and the Tallava dynasties of the Tamil
country. The Dhangar are now, however, a Marathi-speaking community,
hardly to be distinguished from their Kunbi neighbours. The Ilolkar Chief
of Indore belongs to this caste, and still enjoys hereditary grazing rights
in parts of the Dekkan and some of the best of S'ivaji's celebrated
"Mavali" troops were Dhangar. Some of the Dhangar return themselves
as Hatkar, a title for which more than one definition is available. In
some cases, as in the south Dekkan, the Hatkar may by now be a
subcaste, as those who use the name are almost all blanket weavers,
whereas the Dhangar does not always make up his own material. The
derivation of the caste title is uncertain. It has been connected with
"dhan" wealth, or cattle-dealing, an occupation which a few of them
still follow in the south, though most devote themselves to sheep and
goats. In those parts, it should be mentioned in connection with the
above derivation, the peasant habitually refers to his cattle as Dhan, or
Laksmi, that is, the pecunia, or wealth par excellence. The
name of the Kurubar, too, is used for sheep in Kanarcse. In the
south, the caste is called Kurumban. There are two sections; the
pastoral and the Kadu, or jungle, Kurubar. The latter are hunters and
dwellers on the outskirts of the Nilgiri and other forest ranges, and are
still in a very unsettled condition. They are probably the remnants left
behind when the Kurubar of the open countr\' swept down towards the
south-east, and took to cultivation. Even now, the shepherd sections have
elaborate rites and forms of worship alien to those of the rest of the vil-
lagers, and are regarded as not quite assimilated into the community.
The Tamil shepherd is the Idaiyan or those who live by the "middle"
group of the village lands, that is, the pasture. By some of the modern
members of the caste the name is derived from Jadav, connecting them
with Mathura and the Krsna legend. Unfortunately for this tradition, the
names of their subdivisions in some cases connect them with the Paraiyan.
Their present rank, however, is far above that of the latter, and they
are received by respectable castes, in view, it is stated, of their use in
the provision of clarified butter, a pure and popular article in the house-
hold. The Bharvad of Gujarat, belongs, apparently, to the Mcr, one of
the derelicts of a Sc^fthian inroad, which left them in Sindh and the west
of Rajputana. This caste shares with the Khadva Kanbi the peculiarity of
celebrating its marriages only at long intervals, such as 10, 15 or even
20 years. The occasion, as may be reasonably supposed, is one of pro-
longed and uproarious revelry, mingled with elaborate ceremonial, the
details of which are doubtless of considerable ethnological interest. The
Bharvad is also connected with the Rabari, already mentioned as the

I04 5- Ethnography.

camcl-brecdcr of Rajputana. They worship goddesses, especially Mata,
under various manifestations, and have the usual reputation of wanderers
for remarkably potent spells and charms, which ensures them respect.
Nearly all these castes, north and south, are the subject of proverbs
commenting upon the stupidity of their men and the slovenliness or dirt
of their women. The last attribute may be due to the practice of wearing
homespun woollen garments, the durability of which exceeds the means
oF desires of the wearer for purification. In addition to their dealings in
woollen fabrics and, amongst some castes, the provision of sheep and
goats for slaughter, the shepherd earns' a good deal by the sale of the
manure of his flock. In upper India it is the practice to sweep the place
where the latter was penned for the night, and sell the results. In the
south, the utilisation of the product is more complete, and an occupant
pays the shepherd for penning for so many nights upon the sites selected
for the purpose.

§ 71. Earthworkers and Well-sinkers (1,284,3001 Socially speaking,
there is a noteworthy gap between the pastoral castes and the rest of
the nomads, of whom the navvies or earthworkers by profession stand
first. Indeed, except for their dirty habits and their addiction to rats and
other unclean food, these last would occupy the place to which their
skill and industry entitle them. They are practically of one origin under
various titles. In the Dravidian country, where they are most numerous,
they are called Ottan in Tamil, and Vaddar in Telugu and Kanarese.
It is by the name of Od or Odia that they are known north of the Dek-
kan, up to the Panjab. The derivation usually accepted in the south is
from Ofiya, formerly Odra, and now Orissa, as it was from that region
that these gangs are said to have first emanated. Their appearance shows
that they belong to the darker race, and their language, though modified
by distance into a variety of local dialects, has a Telugu basis. In the
south, the Vaddar are generally found in two subdivisions, which do not
eat together or intermarry. The first, and higher section are the Kallu,
or stone quarriers, who are stationary, and abide by their quarries. The
others, called Mannu, or earthy, Vaddar, are migrator)-, and seek jobs upon
large undertakings, working together in their own gangs, by the piece,
in the manipulation of which standard they show marvellous resource and
ingenuity. They are adepts with their large spades, and no unskilled
labour can touch them in the output, either on the flat or in well-sinking.
The ddia reached the Panjab through Rajputana, and seem to have
gradually worked their way up by stages, until they found a supply of
work which maintains them throughout the year. Thus they do not, like
many of the migratory tribes, return to their native country, but settle
in the Province. In the upper parts of the Jamna valley, for instance,
they seem to have given up their traditional pursuit and taken to weaving
coarse cotton wrappers, with a little cultivation thrown in. Here, too, they
have assimilated the local religion, and with the exception of a few details
and ritual, do not keep up their own peculiar customs. Amongst other
refinements, they have raised the standard of their diet, and abjure pork,
one of their favourite meats in the south. In the Panjab a good many
have been converted to Islam, especially those on quarry work. There is
one other caste which shares with the Odia the work of the navvy, viz.
the Bcldar, or the wielder of the Bel, or mattock. This caste too, works
at both stone and earth, and it seems probable that it is a branch of the

Castes and Caste-Groups. E. Nomadic Castes. 103

Odia, detached locally, for the Beldar of Bihar and Oudh has an Od
sub-caste, and also eats rats. In the Panjab, too, the two communities
are considered to be identical, Beldar being merely a functional title. On
the other hand, in Bihar and its neighbourhood it is thought that the
Beldar is a branch of the Nuniya, or saltpetre-maker, which, in turn, is
an offshoot of the labouring caste of the Bind. The BOldar of Bengal
works to a great extent in the coal-mines. Like the Odia, he carries
on his head the earth excavated, and will not degrade himself by putting
the basket upon his back or shoulder. The Kora, on the contrary, his
only rival in this class of work, despises the Beldar for not using the
shoulder-pole and carrying two baskets at once. All the same, the
Beldar holds the higher position and employs a better class of Brahman.
The Kora, or Khaira, a sub-tribe of the Munda race, is closer to his
tribal associations, and the Brahmans who minister to this caste are put
out of communion by their kind. A few other castes have taken to earth-
work as their profession, but they are chiefly small subdivisions of a larger
tribe, such as the Bavariya, who traditionally follow other callings.

§ 72. Knife-Grinders etc. (37,000). There are a few small castes
which may be fairly termed travelling artisans rather than gipsies, since
there is no stigma attached to them personally nor is their calling held
to be a mere cover for criminal means of gain. The Saiqalgar, or S'ikligar,
for example, is a Muslim caste which travels throughout the open season
grinding knives and scissors, and at other times plies in the cities. A
subdivision undertakes the care of razors. In old times the Saiqalgar was
the armourer and polisher of weapons, but he is now in sadly reduced
circumstances. The GhisadF is a small Brahmanic caste of the Dekkan,
corresponding to the Saiqalgar but of lower origin, probably from Gujarat.
The Khiimra is another small Muslim caste of upper India the function
of which is to quarry and sell the querns or millstones for domestic use.
They are hewn at the quarry- and hawked about on pack-animals. The
roughening of the face of the stone after it has been in use a long time
is in Central India and the Dekkan, the work of another caste, the Takari
or Takankar, Brahmanist by faith and nomad by habit. The Khumra's
conduct is above reproach, but the Takari is said to utilise the time he
spends squatting on the premises where he is employed in scrutinising
the extent and disposition of the moveable property of the household,
with a view to a further visit by night, for its removal. The caste is af-
filiated to the great tribe of wandering hunters, called Bavari or Vaghri,
to be mentioned later, and seems to have entered the Dekkan from Gujarat
or Central India, as its members keep aloof from the Pardhi, or hunting
tribes of the south, and speak a dialect resembling Gujarat!.

§ 73. Bamboo-'Workers (295,200). The making of mats, brushes
and weavers' combs is an occupation associated with a gipsy life, not only
in India but wherever these nomadic tribes have established themselves,
and generally connotes an inclinaticm towards burglary or at least petty-
larceny. In the east, moreover, the girls of the castes in question are
usually engaged in ministering to the sexual pleasures of the lower classes
and even of those of the upper who dare to run the risk of excommu-
nication from their caste. There is a more or less definite line drawn,
however, in India between these castes and those, equally low and impure,
who devote themselves exclusively to working in bamboo, a plant which
in several cases has become the totem of the whole tribe, and is wor-

io6 5. Ethnography.

shipped accordingly at the annual caste gatherings. With the exception
of the Turi of Bengal, who are a branch of the great Munrla tribe, most
of the cane-workers of eastern and northern India belong to the Dom.
But, as has been already mentioned, the subdivisions which have taken
to this work are generally settled on the outskirts of villages, not
wandering like the rest, and give themselves the name of Bansphora,
Basor, or otherwise, in token of their profession. In upper India they
admit outsiders into their community after payment of scot and submission
to initiation. In Bengal, the Bansphora are said to be derived from the
Patni, or fishing tribe of the Dom. The Turi just mentioned are practi-
cally a functional branch of the Munda, and keep up their tribal cxogamous
customs and divisions, worshipping the tribal gods under Brahmanic
auspices, and with some regard for Brahmanic precepts as to feeding
with other castes. The Dharkar of the south Ganges valley are also
not far removed from the forest tribe, but have settled round villages, and
employ the Baiga priests, or, at best, the Ojha, a degraded Brahman of
non-Aryan origin. They are considered a much less settled and civilised
community than the Bansphora Dom, but are credited with similar descent.
In the Dekkan and south the Buriid and Mcdar are similar castes, some
of which are settled, others wander during the open season and settle
near villages for the rains. The Medar are chiefly found in the eastern
Telugu districts, and claim to be Oriya by origin. They have subdivisions
which never wander, and are gradually asserting themselves to be Balija,
employing Brahmans and prohibiting their widows from remarrying. The
Burud also are of Telugu or Kanarese origin, and where settled in the
Dekkan are often Lihgayats. In the Tamil country the corresponding caste
is called Vedakkaran, and is probably an offshoot of the northern community.
§ 74. Mat and Basket Makers (348,500). These callings, as just
mentioned, are often, if not usually, the cover for less reputable means
of livelihood, amongst which fortune-telling is one of the more respectable.
Most of them admit recruits from higher castes, a form of accretion which
generally arises from illicit connections with women of the caste, some
of whom appear to be specially attractive even to those far above them
in rank. Thus all the larger bodies are much subdivided, and the general
tie between the communities is very loose. The Kanjar, for instance, 01
upper India, has a section which has never emerged from the jungle or
hunting stage, whilst others never go far from the villages, and make
their living by the manufacture of weavers' brushes, winnowing fans and
the reed-mats used for their own tents and the tilts of the peasants'
waggons during the rains. They also cut querns like the Khumra, and
make leaf-platters like the Bari, and stretch the skins of small animals
for drums. They are said to reserve a certain number of their girls for
marriage within the community and to prostitute the rest. As a rule, they
haunt the Jamna valley and the east Panjab, but gangs are found to
the south, whither they penetrated by way of Central India, and enjoy
a reputation even worse than in the north. As in all castes of this de-
scription, the women enjoy a position of much authority, owing, it is said,
to the frequent absence of their husbands in the seclusion of the district
Jail. If the incarceration be for a long period, a temporary connection
with another member of the caste is formed to bridge the interval. Most
of the castes are Brahmanist of a low type, worshipping the local goddesses,
and not troubling the Brahman. In the south, the great gipsy tribes are

Castes and Caste-Groups. E. Nomadic Castes. 107

the Koraca, Korava or Kuravan and the Yerukala. These used to be
considered identical, and no doubt they come from the same Telugu
stock. They are now separate, however, in both customs and intercourse.
Of the two, the Yerukala, of Telingana, are the more respectable, though
the difference is not great. They have considerable repute as fortune-
tellers in addition to their skill at reed and cane work, but their habit
of travelling with a considerable herd of pack-animals and sometimes pigs,
like the Kanjar, renders them unwelcome visitors in the neighbourhood
of the village crops, which suffer from their depredations. One of the
larger subdivisions of the Koraca derives its title from the carriage of
salt from the coast, and still travels to some extent in that line. They
are superior to the northern tribes in regard to the chastity of their women,
so far as outsiders are concerned, though their facilities for divorce inside
their own body have on several occasions been brought to the notice of
the Civil Courts of the Madras Presidency. The Th6ri of Gujarat are
few in number and probably allied to the Vaghri, a hunting tribe from
the nt)rth. They make and sell bedsteads and mat-work, and live about
in small tents, like the Koraca, using the ass as their means of transport.
The Kaikadi are probably a north Dekkan branch of the Koraca.

S 75- Mimes etc. 148,0001. Owing to the subdivisions of these castes
and the uncertainty as to their origin the figures obtained from the Census
are probably far from accurate. The Bahuriipiya, for instance, or the caste
of many faces, is merely a functional body in the Panjab, and the caste
going by that title is a division of the Mahtam, a hunting caste, which
is said to have got the name from the variety of the ways in which it
picks up its living. In the Ganges valley, on the other hand, the Bahurupiya
is a sub-caste of the Banjara, and takes brides from the Nat, another
gipsy tribe, but gives none in return. The Mahtam too, are connected
with the Labana of the Panjab, so it is not unlikely that the Bahurupiya are
really of the latter blood. This caste stands much higher than the Bhand,
or Buffoon, who plies his trade about the mansions of the great, like the
jesters of old, and with even greater freedom of speech. Indeed, the ill-
temper of the Bhand is proverbial, mainly because of the peculiarly offen-
sive manner in which he gives vent to it. In the Panjab the caste is
recruited largely from the Mirasi, whose name is sometimes retained as
well as that of the trade. The Bhavaio of Gujarat, is an acting caste,
and performs comedies at weddings or other festivals before any village
audience subscribing for it. The company is often attached to the village,
as part of the establishment. They have the tradition of having once held
a higher position in the north, but are now a purely local institution, and
owing to confusion of nomenclature, perhaps, their full strength has not
been recorded. The GondhaH of the Maratha country is an itinerant ballad-
singer, and dances a special set of figures in honour of a goddess at
weddings and private entertainments.

{; 76. Drummers (206,200). The ceremonial drummer of a village or
temple has been referred to as usually belonging to one of the resident low
castes, and is generally upon the village staff. There are others, however,
who are more strictly professional upon this instrument, and wander
about for their living. The Dafali, for instance, and the Nagarci, of
the Ganges valley, are ISIuslim, with a sort of religious flavour about their
performances. The former expel spirits as well as extorting alms. The
Dholi of Rajputana, like the Bajania of Gujarat, are Brahmanist functional

io8 5. Ethnography.

castes, recruited from the village menial and scavenging classes. The
Turaha blow horns and are only found in Bengal.

§ 77. Jugglers and Acrobats etc. (235,8001. There arc numerous
bodies of jugglers, tumblers, snake-charmers and the like, each with a
different name, but all connected, at least in upper India, under the ge-
neral title of Nat or Bazigar. It is difficult to say how far the former
is the designation of a caste or of a function. In the Panjab, for instance,
Nat is usually held to be a caste, and Bazigar the branch of it which
takes to juggling and tumbling. In the Gangetic region, again, the Bazigar
is a subdivision of the Nat, like Badi, Sapcra, Kabutara, denoting different
performances. Then, in Bengal, the Nat or Nar is a caste of trained
musicians and dancers of much higher position and accomplishments, and
quite distinct from the nomad of the same name. Further to the south,
there arc the Dombar or Dommara, of the Tclugu country, who are
identical with the Kolhatf of the Dekkan, both sharing the occupations
and traditions of the Nat of the north. In addition to their acrobatic and
similar performances, the greater portion of these communities live by the
manufacture of horn articles, by hunting the wild pig and by prostituting
their women. They hold themselves above the pom and village tanner,
but almost invariably feed on vermin or carrion. Except in the Panjab,
their appearance is that of the dark races of the Central Belt, and,
indeed, a good many of the clans say that their original home was
amongst the Gond tribes of the eastern parts of the Central Provinces.
There are, necessarily, different grades amongst them and the distinc-
tions are strictly maintained, but most will admit members of higher
castes upon payment of a caste-feast or other means of establishing
a footing. They are not by any means all criminal, though most are
credited with the propensity- to break into houses and steal fowls and
cattle when the opportunity occurs. The small section of the Gopal, for
instance, of the Dekkan, is a notorious cattle-lifter. In some of the sub-
castes of Nat only the men perform. In others the women are kept for the
tribe, and do not prostitute themselves to outsiders. This, however, is
exceptional. In one of the sections, the women are experts in tattooing,
and act as professionals in this art for other castes, as the Koraca do
in the south. About three fourths of the Nat are Brahmanists of a low type,
with their own special deities and forms of worship. Occasionally they
obtain the good offices of Brahmans, if only to fix the lucky day for their
ceremonies. Their jungle origin is indicated in a good many cases by
their knowledge of roots and herbs as medicines, together with their pos-
session of secret preparations of repute as aphrodisiacs, love-philters and
the means of procuring abortion, for all of which there is a certain and
constant demand amongst the better classes.

§ 78. Thieves (133,500). Along w-ith the above may be taken the
castes which have little or no means of livelihood except stealing. In
some cases this general condemnation must be qualified, as the same
caste may be criminal in one locality but innocent in another. The Ba-
variya, for instance, is simply a fowling caste in the Panjab, where it is
most numerous in that capacity ; but the Bavari or Bagariya of Central
India and the north Dekkan, where it has several sub-titles, is always
under the eye of the police during its travels. The Bediya, again, bears
a very bad character along the Jamna and in Oudh, but has quite re-
spectable sub-castes in Bengal, where many have accepted Islam. Another

Castes and Caste-Groups. E. Nomadic Castes. 109

sub-caste, however, the Mai, is closely connected with the K61 race, and is
credited, indeed, with the parentage of the whole Bediya community. In upper
India that relationship is obscured if not contradicted by the affinity of all
these castes, such as the Bediya, Habura and the like, with the Sahsiya,
the thief par excellence, of the north. The exploits of the last-named
community have given it a celebrity which is not justified by its numerical
strength, though owing to its subdivisions it is difficult to ascertain the latter.
The Sansiya stands in curious relationship to the Jat tribe, each family of
which has its Saiisi genealogist. When a question arises in connection
with pedigree it is said that the word of the Sarisi is accepted in pre-
ference to that of the IMirasi. It is not easy to trace the origin of this
parasitic attachment o( the degraded caste to the undoubtedly pure and
foreign body, especially as this is the only function of the Sansiya which
does not bring the caste into unfriendly contact with the police. The
women, no doubt, sell roots and herbs, but their object in so doing is
said to be merely to get access to the inside of the domicile, and thus
obtain information conducive to burglary by their husbands. In contra-
distinction to the practice of the Nat, the Sarisiya women are said to be
chaste in their relations with outsiders, like the European gipsies, and
very staunch in their defence of their male relatives when trouble is
imminent. They thus enjoy much influence in the tribal councils, and,
owing to the natural timidity- of the caste in applying for the protection
of the law, these councils practically regulate all the affairs and disputes
of the community-. It is hardly necessary to say that their religion is of
the most simple, and that they feel bound to call in outside spiritual aid
only in cases where the ghost or demon of the locality has caused serious
illness or bad luck. A few of them have been converted to Islam, but
one large section asserts its Rajput origin and keeps aloof from the rest
of the tribe. On the other hand, it has been found advisable to form a
subdivision to meet the case of the half-breeds, sprung from outsiders
who have been admitted into Sahsiship, generally owing to devotion to
a girl of the tribe. The small caste of the Habura, along the upper
Ganges and Jamna, is allied to the so-called Rajput section of the Sansiya,
and keeps up regular Rajput sept divisions. It resembles the parent tribe
in its care of the women and disregard of the rights of property', but it
seems to be rather more Brahmanised in its customs and is less given
to crimes of violence. In the thieves' latin of all these criminal tribes of
the north, it is interesting to trace the strong element of corrupt Gujarati
found throughout, and the same feature is noticeable in the slang of the
north Dekkan tribes of this class, as if the western Vindhya had been
the nucleus of errant criminality among the Kol races. In the Dekkan
itself and the Karnatic, the only tribes of this class are the small com-
munities of Bhamtiya, Ucli or Ganticor, habitual pilferers, but not further
advanced in crime. They are settled in some strength in Poona and its
neighbourhood, where their calling has proved so lucrative that several
have become large landholders. The railway has been the making of them,
as they travel in disguise over the length and breadth of the country,
cutting purses and slitting up bundles and carpet bags on their way. They
are of Telingana origin, and still keep up their worship of Yellama, the
Earth-goddess, of their home. The Sanaurhiya, another travelling frater-
nity of the same pursuits, do not appear in the Census returns, since they
return themselves as Sanadh Brahmans. They are a composite community

no 5. Ethnography.

recruited from all sorts of castes, but now bound together by the usual
caste regulations, including one prohibiting all crimes of violence. Their
head-(|uarters are in Bundclkhani.!, but they are mostly on the move in
disguise, with a few of their more wealthy members established in the
chief towns to act as receivers of the goods obtained on the journey.
Herein they differ from the Sansiya, who will not venture into the town,
but concert a meeting in the open field with a Sonar or other respectable
member of society, with whom the bargain is made, and the goods de-
livered accordingly.

§ 79. Hunters and Fowlers (977,600). This is a group which in
one direction is merged in that of the lower cultivators and field-labourers,
and in the other undoubtedly tends towards that of the petty criminal.
The same caste may have a branch in one province entirely devoted to
settled village life, whilst in another part of the country it is still in the
jungle or nomadic stage. So far as upper India is concerned, there seems
reason to think that most of the hunting castes of the present day take
their origin amongst the dark race of the western Vindhya. Their own
traditions point, as a rule, to north Rajputana as their native country, but
as the south is approached, the hills of Malva and the west assert their
influence, and relationship to the Bhil or other Kol tribe is claimed.
Several of the tribes take their name from some implement of their trade,
usually the net or noose, as in the case of the Vaghri, Valaiyan and
Bavariya, and the Phansi-Pardhi, of the west, without any indication of
their parentage. The Bavariya is a particularly varied community-. It has
all the appearance of Kol descent, even in the Panjab, where it has long
been established. Here the caste is said to have come from Mevad and
Ajmer. It is subdivided into three sections, only one of which still gets
its living by the noose. Of the rest, one has taken to cultivation, and
the other to vagrancy and petty crime. They are all by heredity good
trackers, and though foul in their diet, not badly looked upon by their
neighbours when they are settled. Along the Jamna, however, their
character deteriorates, or more correctly perhaps, has not yet risen to the
level it reaches further from its native haunts. It is, however, fairly well
Brahmanised, though it keeps to its own worship. The higher castes are,
as usual, admitted on payment of the cost of a feast, or even by eating
with the members of the tribe. One of the subdivisions, the Moghiya, is
often considered a separate caste, but it seems to be no more than
the Central Indian variety of the main body. The Bavariya of the eastern
parts of the upper Ganges valley are apparently quite distinct. They assert
Rajput origin and came from Baisvara, and employ the Panrc Brahman
of their former residence. In spite of their dark complexion and non-.-\ryan
appearance generally they are not connected by their neighbours with
any of the local hill-tribes, and are received on terms of equality by the
peasantry and others. The Ahi-riya, a tribe found both in the Panjab and
along the Jamna, is similarly divided. In the north they are hunters and
reed-workers and occasionally settle down to life in connection with, but
outside, the village community, without any suspicion of criminal tendencies.
Along the Jamna, however, their reputation is that of potential burglars
under the guise of mat-makers and collectors of jungle produce. They
were formerly renowned for the well-planned gang-robberies they effected
at long distances from their homes, and like the Bhils, for the expedition
with which a large body could be got together from many different quarters.

Castes and Caste-Groups. E. Nomadic Castes.

and melt away imperceptibly as soon as its purpose was served. In the
present day, they use the railway, and organise expeditions far away in
Bengal and the Panjab. The caste is peculiar in having no subdivisions,
endogamous or exogamous, and the conversion of one of its members
to Islam makes no difference in his social position. The Bahcliyi is
another example of the same name being borne by separate communities.
In Bengal, the caste is said to be akin to the BediyS, mentioned above,
and is almost exclusively occupied in hunting and fowling. In Bihar, the
BaheliyS, or Bhula, is called a sub-caste of the DOsadh, but will not hold
social intercourse with the latter. In the Ganges valley, again, this caste
is said to belong to the Pasi, whilst in the west, it is affiliated to the
Bhil, and is claimed as kin by the Ahcriya. In spite of their occupation
of fowling, they are not amongst the impure, and though unattached to
most of the ordinary Brahmanic forms of worship, they observe the orthodox
festivals and employ the village Brahman for their own sacrifices. Com-
paratively few of them are Muslim. So many are now resident in villages
that they are no longer to be counted amongst the nomad tribes. The
same may be said of the M ah tarn, a hunting caste of the Panjab, chiefly
found in the Satlaj valley. Only a section of them still live by their
traditional use of the noose, and the others are settled cultivators and
labourers, with a good reputation for industry and quiet behaviour. Portions
of both sections have changed their religion to Islam or the Sikh creed,
but preserve withal much of their original habits. There is another com-
munity of the same name in the submontane tract of the Panjab, which
seems to be a branch of the Banjara or Labana caste, and to have made
its way from the east, whereas the hunting Mahtam reached the Satlaj
from Rajputana. There is thus no connection between the two. One other
caste of the Vindhya belongs to this group, namely the Sahariya, ot
Bundelkhand and the neighbourhood. It is said to derive its title from
the Savara, a name now reserved to a tribe of the south Orissa hills, but
applied by Sanskrit writers to any of the Dasyu tribes of the Central Belt.
Beyond a common darkness of colour and similarity in feature, there is
no link between the two traceable in the present day. The Sahariya do
not wander about the country more than is necessary to give them a good
supply of the jungle produce which they live by selling, and their crimi-
nality is confined to petty thefts and an occasional gang-robbery. The
caste seems to be subdivided on totemistic lines into a number of e.xo-
gamous sections. They profess Brahmanism, but worship chiefly their local
demons without the intervention ofBrahmans. There is no tradition amongst
them of having immigrated from any other part of the country. The other
side of the Vindhya presents but few hunting tribes, and those mostly of
northern origin. The Vaghri of Gujarat, who are apparently the Baghri
of Central India, say that they are kinsfolk of the Sansiya of the Panjab,
and came from north Rajputana. They are now, however, naturalised in
the west. In that part of the country they are subdivided according to
function, and, where they are numerous, according to geographical sections
which do not intermarry. They are still great hunters and bird-snarers.
In the latter capacity, they have struck out a new and lucrative line of
business with the Jain and other Vania, who set a very high value upon
animal life. The Vaghri makes his catch of birds, takes them in cages
to the house of the trader, and there offers to kill them or let them be
ransomed, knowing that the merit to be acquired by the latter process

;. Ethnography.

will outweigh the cost in the mind of the orthodox. They also keep fowls,
and rent fruit and other productive trees by the year, selling the crop.
Most of them wander during the fair season, but a good many have settled
down near villages. They have their own priests or clan-elders ("Bhuva),
who perform their ceremonies and regulate the caste generally. The
Vaghri, though not quite in the ranks of the criminal castes, has a bad
reputation among villagers for theft. In the north Uckkan, indeed, this
caste is credited with a good deal of the crime against property, but it
is not certain that the sub-castes which operate in that region are not
from Central India. Linguistic evidence seftms to indicate a Gujarati origin,
but, as stated above, this peculiarity is found in the dialects of tribes far
separated from that province. The Phansi-Pardhi, however, or snarers
of bird and beast, seem to be really a branch of the V'aghri who have
made their home in the Marathii country, where they are occasionally
found in the capacity of village watchmen.

Up to a certain point all the hunting castes in the Dekkan assert
their origin to have been in the north. After that, the corresponding castes
claim to have come up from the south. The Berad or Bedar have been
classed with the watchmen, and so have the Tamil castes now so engaged :
but there seems reason to think that all these castes are connected in
some way or another with the Vedan, Valaiy an, Vettuvan and similar
bodies, the majority of which belong to the hunting or fowling order.
What the connection really is has not yet been ascertained. There is,
however, a sub-caste of Ambalakkaran bearing the name of Vcdan, and
the whole body claims to be descended from a Vedan, and the Valaiyan
say that this same hero was the founder of their caste also. The Vettuvan
hold their heads higher, and add the title VeUalan to their caste-name,
saying that they were imported by the Kongu Chiefs to assist them in
the conquest of Kerala. The Vedan say they were originally natives of
Ceylon, and the Vettuvan worship Kandi-amman, the goddess of Kandy,
as well as their seven Kannimar, or tribal deities, worshipped also by
the Irula, a more primitive tribe. The Vettuvan of the interior, again, are
distinct from the caste in Malabar bearing the same title. Another small
hunting caste in Malabar is the Kuriccan, confined chiefly to the Vainad.
The former stand higher than the latter, though both are jungle-haunters.
The Kuriccan, too, have the same abhorrence of contact with the Brahman
that the Paraiyan have, and worship a tribal god of their own. It would
seem, therefore, that excej)! in the west, these castes are more settled
and likely to rise in position than any of those found in the north, and
that the members or families which continue to follow the traditional oc-
cupation are being gradually relegated to sub-castes below the general
level of the rest.

F. Hill Tribes.

§ So. It can be easily inferred from what has been set forth in the course
of this survey that the importance in the ethnology of India of the pre-
Aryan inhabitants can scarcely be overrated. There is, on the one hand,
the gradual extension among them of the foreign forms of speech; on
the other, the assimilation of their forms of belief into the religious system
of those who have dispossessed them of their territory and position. In
the preceding portion of this work, too, instances are given over and over

Castes and Caste-Groups. F. Hill Tribes. 113

again of the incorporation of communities, wholly or in part, into the
Brahmanic social system, showing the extent to which that system and
the racial constitution of the population at large is permeated from toj)
to bottom by the Dasyu element. It becomes necessary therefore, to give
some consideration to the remnants of these primitive communities which
have, so far, more or less escaped absorption, and have preserved in a
modified but still distinguishable, shape their independent tribal existence.
It is obvious that in the present day the chief interest of these tribes is
found, cthnographically speaking, in their constitution, customs and beliefs.
Into these subjects it is impossible to enter in the detail they merit in a
review of this description. It is also unnecessary, as they have been treated
for the most part by experts, in works devoted to such investigation, and
the rest are still the subject of inquiry in similarly competent hands. All
that is here attempted is a cursory sketch of the position, strength and
geographical distribution of the more representative of these bodies, in
order that their place in the Indian Kosmos may be duly appreciated.

It is convenient to treat of these tribes according to the tracts wTiich
they inhabit. The most important of these, in both extent and ethno-
graphical interest, is what has been called in this work, the Central Belt.
It comprises the great plateau of Cutia Nagpur, with an extension to the
north across the Santal Parganas to the Ganges at Rajmahal. Southwards,
it follows the ranges which separate Orissa from the eastern parts of the
Central Provinces, skirting the plain of Chattisgarh, and continuing south
as far as the lower Godavari. Westwards from Ciitia Nagpur, the hill
country passes along the south of Shahabad and Mirzapur, along the Kaimiir
range and the Vindhya, to Mcvar and the Aravalli. Almost parallel, to
the south of the Narbada, are the Mahadcv and Satpura ranges of Bcrar
and Khandesh, ending in the forests of east Gujarat. Contiguous to this
western abutment of the Belt, is the line of the Sahyadri, or Western
Ghats, which, about as far as the little State of BhOr, is inhabited by
a few small tribes of the same character as those further east, and pro-
bably allied to them in race. Then there occurs a gap in the series, as
the south Dekkan is cultivated almost up to the edge of the Ghats; and
the next locality in which the more primitive tribes are found is the
Nilgiri, with their detached continuation separating Travancore from the
east coast. The above tracts are the present homes of the remains of the
Kol and Dravidian tribes. The hill communities of Mongoloidic race are
found chiefly in the ranges separating Assam from Upper Burma, and in
the dorsal range of Assam itself, made up of the Garo, Khasia, Jaintya,
Naga and Mikir hills, between the Brahmaputra valley and the Deltaic
plain. The remaining group inhabit the Himalayan southern ranges, and,
being chiefly resident in Nepal and Bhutan, countries beyond the census
limits, come but slightly within the scope of this review.

§ 81. (a) Central Belt (9,221,900). The tribes of this tract may be
taken first, not only because they form the largest division, but also by
reason of their more intimate racial connection with the masses of the
plains. Each differs from the rest in some important respects with regard
to organisation, customs and beliefs, but there are a few characteristics
general throughout the whole. All but three or four of the larger tribes
believe themselves to be autochthonous, if not to the tract they now in-
habit, at least to one within a comparatively short distance. All the larger
tribes, again, have traditions of dominion over a much larger tract than

Indo-Aryan Research. II. 5. 8

114 5- Ethnography.

their present one, and in most cases the statement is supported t)y evidence
such as that of ruins, names of places and castes and by identical forms
and objects of worship. In every large tribe, again, there are sections
which are far more Brahmanised than the rest, usually with the tendency
to separate under a different title, the latter being borrowed from an
orthodox community of the plains. Most of the tribes arc much subdivided
into exogamous divisions, totemistic as a rule; endogamous sections following
later, after contact with Brahmanical castes. Where the tribe is free from
such outside influences it employs priests belonging to its own or a neigh-
bouring community, and in several cas6s the more imjiortant sacrifices
are performable in the archaic fashion by the head of the family only.
The usual form of religion is that of the worship of nature or spirits,
with the accomi)animcnt of spells, witchcraft and exorcism generally.
Among the more Brahmanised tribes there is the outward acceptance of
some manifestation or other of a member of the Puranic pantheon, but
from the practical side of devotion and propitiation, the belief in the
efficacy of the older system remains unshaken: and it is worth noting that
the older the tribe in the locality the higher the reputation of the priests
it furnishes. In regard to occupation, the greater part of this population
lives by cultivation, a few tribes on the plateau of Cutia Nagpur having
attained to a fair degree of skill in their calling, and making use of the
plough. The bulk, however, still pursue the primitive and wasteful system
of clearing a patch of jungle, burning the vegetation thereof for manure,
and raising two or three years' harvest off it. They then leave it fallow for
some years, moving off meanwhile to another patch. Where this is the
practice, the village is migratory, within a certain range, or consists merely
of detached hamlets; but in the more open country, cultivation being
permanent, the village site is so too, and the huts or houses are built
more solidly. In a few of the wilder tribes the whole village is apt to
flit when untoward events have proved the locality or its deities to be
unpropitious. In all the large tribes there are sections which live almost
entirely upon forest produce, and in some, where an autumn crop only
is raised, the people rely during the hot weather entirely upon what the
jungle contains. Some communities, again, make it their regular trade to
collect lac, tussar-cocoons, berries and other produce for sale to agents
from the towns, whilst others habitually work in cane or make tooth-sticks
and brushes, smelt iron, or wash the river sands for its minute yield of
gold. At the other end of the scale arc found in several tribes landed
proprietors of considerable wealth, who have long passed out of the tribal
into the caste stage, and who, in the case of petty Chieftains, marry into
respectable Rajput families — at a distance. Great advantage has been
taken by others of the opportunities of earning good wages on the tea-
gardens of Darjiling, the Tarai and Assam, where they bear an excellent
character for industry and docility. The heart of the jungle, however, has
hitherto proved almost imi)ervious to the efforts made to improve the in-
habitants by land grants or other means of inducement to them to work
themselves into a higher material condition. On the outskirts it is different,
and there, as before remarked, the tribal population is breaking away
from its traditions, and becoming merged gradually into the conditions
of the plains.

§ 82. It was pointed out in the introduction that whilst in physical
characteristics and general customs these tribes appear homogeneous.

I

Castes and Caste-Groups. F. Hill Tribes. 115

in speech they fall into two different categories, the Kol and the Dravi-
dian. In the case of most of the southern tribes this distinction is ob-
viously attributable to the contiguity of the Andhra or Telugu population
of which they form the northern fringe. As regards the detached com-
munities further north, however, there are traditions of immigration,
and it is remarkable to find tribes like the Oraon, of the south of Cutia
Nagpur, and the Mal-Pahariya and their neighbours of the hills bordering
the Ganges speaking tongues which support their assertion that they
reached their present localities from a tract as far distant as the Karnatic,
especially when to do so they must apparently have outflanked the Gond,
a still more powerful tribe, which itself is said to have come from the
same home. It must be noted that the Male, or n<irthern section of the
inhabitants of the Rajmahal hills, are also called Savariya, or Sabar, a
title which appears to link them to the Savara, or Suari, of the ancient
European geographers, Pliny and Ptolemy. These were once undoubtedly
in possession of a considerable territory south of the Ganges, but now
the only large tribe known by their special designation except the Brah-
manised Sahariya, mentioned above, is located far to the south, and isolated
amongst a jiopulation speaking either Oriya or the hill-vernaculars of the
Dravidian type. On linguistic grounds, the Savara of today arc grouped
amongst the Kol-Khervari peoples, whereas the INIale use a tongue nearly
akin to that of the Oraon. It is possible, therefore, that an ancient and
wide-spread title has been applied to two different and distinct communities,
and that the southern Savara like their neighbours, the Gadaba, are Dra-
vidian by race, modified by the influence of more powerful alien sur-
roundings. Thus, it may be generally put that the Uravidian element is
indigenous in the south-east, immigrant in the south, centre and a portion
of the north-east; and that the north, west, and most of the plateau, ap-
pertain to the Kol-Khervari tribes.

In regard to the latter, it must be noted that the generic designation
of Kol is not returned as the title of a tribe except in the Central Pro-
vinces, Central India, and the south of the Ganges valley. Towards the
east of the tract in question, the terms used are Ho, Munda and Bhumij.
Of these, Ho is held to mean Man, the name given to themselves by most
primitive tribes. Kol is probably derived from Ho by transliteration. Munda
and Bhumij are terms of Sanskrit origin, the former meaning a headman
of a village, also a common appellation for the lower races in India, and
in this case adopted by the tribe itself. Bhumij, in the same way, implies
connection with the soil, and connotes in most cases in which it is applied
the clearers of the village-site. In various forms it is found from Gujarat
to Assam. Occasionally it means the hereditary landholders of the village;
elsewhere, the menials and guardians of the boundaries. In the form of
Bhiiinya, in Bengal, it is both a generic title, covering a considerable
number of castes of different standing and origin, and also the name of
a loose and scattered tribe in the south-eastern part of the Belt. The
tribe to which the name of Bhumij is now given is a branch of the
Munda which has spread from the central home of the race to the east-
wards, and now lives in western Bengal and the districts of Manbhiim
and Singhbhiim. The community is almost entirely Brahmaniscd, except
in the tracts immediately adjoining the plateau, where the Murula language
is still current, and the people intermarry with the Munda of the uplands,
and often call themselves by their name. As the tribe advanced into the

Ii6 5. Ethnography.

plain all this was changed. The tribal worship was abandoned by the
landholding class in favour of Brahmanism of a somewhat strict type, and
the Aryan vernacular of the district is used by them. In the wealthier
families the practice is growing of calling themselves Rajputs and dropping
their ancestral connection altogether. The less advanced adhere to their
tribal gods and employ their own Laya, or priests, on all occasions. The
Munc.la are subdivided into numerous tribes, the names of most of which
prove an origin from intermarriage with other tribes of the vicinity. These,
again, are further parcelled out respectively into totemistic sections,
of course exogamous, and with interestfng rules as to prohibited food.
The chief object of worship is the Sun, as is the case with most of the
larger tribes of this tract, but a more efficient and active deity is found
in the Mountain god, again a not uncommon feature of the K6I race. The
priests, or Pahan, are members of the tribe. The Ho, sometimes called
the Larkha K61, arc probably the oldest, as they are the highest, of the
three cognate tribes. The Santal, Bhiimij and the Munda call themselves
Ho, but no one else does, and intermarriage between them and the Ho
of Singbhum is now unusual. The latter are of Cutia Nagpur, like the
others, but having got possession of a more fertile region, they have taken
the greatest care to prevent strangers from sharing the land with them.
Physically, they are the finest of the race, and have become a steady
agricultural community of a somewhat undeveloped type. The tribes re-
turning themselves as Kol are found for the most part in the Mirzapur
district along the Ganges, in Jabalpur and Mandla in the Central Provinces,
and in the Baghclkhand tract of Central India. They have the tradition
of having once lived in the plains of south Bihar from which they were
expelled by Savara of some sort, and had to take refuge in Baghclkhand.
In all the above tracts the tribe is comparatively Brahmanised and has
lost much of the organisation and worship it has retained in Cutia Nagpur,
where the Kol is a branch of the ^lunda. Here they live after the fashion
of their ancestors, but in the rest of their settlements they have taken to
simple cultivation on the ordinary lines, and differ but little from their
Brahmanic neighbours except in more extended respect for sorcery, and
in the propitiation of the local gods in preference to those of wider fame.
One of the most civilised tribes of this group is the Kharvar, to which
belong more than one of the local Chiefs who have been accepted as
equals by Rajputs, on payment, however, of unusually heavy dowries. The
Kharvar appear to be without traditions of immigration from further than
the south east of the Cutia Nagpur table-land, from which they spread
northwards and down into south Bihar. Here their rank seems to depend
much upon their connection with the land. Those who hold large estates
claim to be Rajputs, and the middle classes employ S'akadvipi Brahmans
and retain only the more important of their tribal ceremonies. Even
amongst these classes the influence of the Baiga, or tribal priest, is by
no means extinct. Indeed, the reputation of the tribe for supernatural
powers is such that a section bears the name of Baiga, and is so re-
turned in the Central Provinces and Baghclkhand, from which it may be
inferred that the Kharvar is regarded by its neighbours in that direction
as being of an older stock than themselves. On the other hand, sections
of the Kharvar now employ a priest of the Korva, or even a lower tribe.
The respect shown by the Kharvar for the Khar grass, which they say
they take their name, seems to indicate that they were once a totemistic

Castes and Caste-Groups. F. Hill Tribes. 117

branch of a larger community, but no traces of this have been ascertained,
and the tribe holds itself to be superior to all around it, except, perhaps
the Ccru. The latter are even more thoroughly Brahmanised than the
Kharvar, and have the same tradition of having been ousted from dominion
in the south of Bihar. They were the last to leave the plain for the plateau,
and are accepted as an orthodox Brahmanic caste. A small section, however,
in the interior, still keeps to the jungle and breeds tussar moths, for doing
which they are deemed impure by their relatives. Long periods of settled
life, combined with frequent intermarriage with high class families of
Rajputs and others, have in fact made the larger body of the Ceru a
different and distinct community, claiming the name of Cohan-bansi. The
totemistic subdivisions of their poor relations, however, prove their con-
nection with both the general Muncja race and perhaps more especially,
with the Kharia. These last say they came up to Wanbhum and Ranci
from the Orissa State of Mayurbhanj. One branch took to cultivation and
settled life, whilst those in Manbhum remain amongst the most shy and
uncivilised of their kind. The former affect the highest regard for purity
in diet, and greatly restrict their intercourse with outsiders, a habit which
is sometimes unkindly attributed to their own filth and disregard of social
decency. They intermarry with the Munda on unequal terms, the larger
tribe taking brides from them but giving none in return. The Kharia keep
to their own worship, using Munijfi or Oraon priests. The jungle section
live on the produce of the forest with a little simple cultivation of the
migratory sort. When any stranger settles within sight, they move off, a
tendency welcomed by their neighbours, who regard them as the possessors
of excepti(mal powers of magic, available against both man and beast.
The largest of the Kol communities is the Santal, who call themselves,
like the Munda, by a term signifying Headman of a village (Manjhi). The
tribe is not autochthonous in its present locality, though their immigration
does not seem to have been from a greater distance than the south-cast
of the Cutia Nagpur plateau. From thence they spread eastwards and
northwards in succession, and peopled the Santal Parganas about the
middle of the 19th century. This eastward movement is still in progress,
and the Santal are gradually taking up land in that direction wherever
they find they can keep on laterite soil and within the range of the Sal
tree, which is said to be to them all that the bamboo is to the inhabitant
of the plains. The aversion from alluvial soil manifested by all the tribe,
is accounted for, according to some, by its unsuitability to their favourite
tree, whilst others attribute it to the fact that the uplands afford better
outlets for expansion of cultivation than the already well-peopled riparian
tracts of the great valley. The Santal is also one of the people most
willing to leave his home for temporary engagements on the tea-gardens
of Assam and the Tarai, where over 40,000 of this tribe were returned
at the Census. In spite of their wanderings, the Santal have kept up their
elaborate tribal organisation, with a most intricate subdivision of clans and
with mystic pass-words current amongst them. Their tribal worship of the
Sun and ^lountain, too, is strictly maintained. Each family, moreover, has
its own domestic god with the addition of a secret god, the name of
which is kept a mystery to the women of the household, and only divulged
to the eldest son of the house, lest undue influence be brought to bear
upon it. It is said that a generation or two ago, the wealtier Santals, in
imitation of the Brahmanic high castes of the neighbourhood, took to

n8 5. Ethnography.

marrying off their girls at a very early age- This practice is common
enough amongst the aspiring families of the lower classes, but the re-
markable feature in the new departure among the Santals is that after a
few years' trial the practice was abandoned and the tribal custom of
marriage in the teens was resumed. There have been a good many
converts to Christianity from the tribe of late years, and, indeed, most
of the information available about the language and religion of the tribe
is derived from Danish and other Missionaries working amongst them:
In their own worship and in the periodical great sacrifices the Santal
relies upon the Naiki, or priest of his own community. Akin to the Santal
is a small tribe called Mahilf, which, judging from the names of its sub-
divisions, must have split off from the main body on taking to work, such
as carrying loads and making baskets, deemed degrading by the Santal.
It seems, too, that the Munda contributed a section to the Mahili. The
latter are now found chiefiy in Manbhum and the Ranci district of the
plateau, with a few scattered amongst their kinsfolk elsewhere. Their
religion has been described as a mixture of "Animism half-forgotten and
Brahmanism half-understood ". Sacrifices are offered to the god of the
mountain and to the snake and then consumed by those who make the
offering. One subdivision only has advanced well into the religion of the
plains, and employs Brahmans and abjures the food dear to the rest. The
Binjhia and Birjia have usually been considered to be one tribe, but
at the last Census it was considered better to tabulate them separately.
This course appears to have been correct, as the larger community of
the Binjhia is a Brahmanised cultivating caste, speaking Ofiya, and settled
in the south of the Ranci district, whilst the Birjia are residents of the
uncleared forest, where they live from hand to mouth by the cultivation
of small patches, eked out by hunting wild animals and collecting fruit etc.
They are held to belong to the Agaria, or iron-smelting tribe whose
customs they follow. The Juang, or Patua, are perhaps the most primi-
tive of all their group. They inhabit the recesses of the Orissa hills, and
it is remarkable to find the caste amongst the indentured labourers in
Assam. Both language and customs indicate their close relationship with
the Kharia and Munda. They worship the forest and village gods, but
are said to be acquiring some appreciation of Brahmanic deities. They
keep village priests, but the important offices are performed by the elders
of the tribe. The latter, probably because it is so small, is not subdivided,
but forms a single endogamous community. The practice of clothing them-
selves with leaves, which has been picturesquely described by Dalton and
other visitors to their haunts, is said to be yielding to the taste for cotton
wrappers, even amongst the women, who have hitherto alleged divine
warrant for the leaf-apron.

§ 83. Of the immigrant tribes of the plateau, the most important is
theOraor, or Kurukh, which, as stated above, is apparently of Kanarese
origin. According to the tribal tradition, the Oraon once held a good
portion of South Bihar, and on being expelled by the Muslim, separated
into two branches, one following the Ganges to the Rajmahal jungle, the
other going up the Son and occupying the north-west corner of Ciitia
Nagpur. The main body are now settled in the latter tract, covering the
districts of Ranci and Palamau. As they are greatly in request as labourers
they are also found in the Census returns of Assam and the Jalpaiguri
tea districts in considerable numbers. Having dwelt side by side with the

Castes and Caste-Groups. F. Hill Tribes. 1 19

Munda for many generations, they have dropped a good many of their
own customs and adopted those of the indigenous tribes. In regard to
their worship, however, they keep themselves apart, erecting some symbol
of their gods, whilst the Munda abstain from anything of the sort. The
Oraon employ no Brahmans of course, and their priests are Naya, very
like those of the Munda. According to tradition, the Oraon introduced the
plough into the plateau and were the first to take to regular cultivation.
They regard the Munda as their predecessors, however, and where the
two are in the same place, the Oraon yield precedence to the elder tribe.
The advance of settled government and systematic land administration
has not conduced to the prosperity of the Oraon, who lose ground before
the more cunning castes which follow those symptoms of civilisation, and
prey upon the less educated, gradually dispossessing them of their lands.
As to the other branch of the Oraon, who are still entrenched in the hills
of Rajmahal, it appears that two sections have been formed, one, of the
Mal-Paharia, the lower and more Brahmanised community, and the other,
called, for want of a more definite title, the Male, or Hillmen. There
seems to be little doubt but that in spite of the antagonism between the
two in the present day they belong to the same race, using closely allied
dialects of the Oraon-Kanarese language. The Southern community, though
more civilised than the Northern, is still more or less in the jungle stage,
and worships the Sun, Earth and Tiger, through the mediation of the
headmen of the villages. One subdivision is considered by the outside
world to be a trifle purer than the rest, as in the matter of diet it draws
the line above rats and lizards, which enter into the daily meal of the
others. They cultivate on the wasteful system of jungle-burning, which
entails the occupation of an abnormally large tract of land to allow of
the frequent fallows necessary for the recuperation of the vegetation. The
Male of the upper hills, are far less affected by Brahmanic contact than
the others, and are said to be homogeneous to the extent of not having
even exogamous subdivisions. They share with the Mal-Pahafia the worship
of the Sun, but differ from the latter in setting up a post to symbolise
that luminary. The only semblance of a priest amongst them is the Demano
or Diviner, and even he gives place to the headman at the more important
ceremonies. The Male gave a good deal of trouble in the early days ot
British rule in Bengal, as they had managed to preserve their independence
of all government against the attempts of the Muslim to coerce them.
The judicious handling of them by a popular local official, late in the
iSth century, pacified them into the abstention of raids upon their neigh-
bours, but his attempts at inducing the tribe to take to industrial pursuits
were not successful.

§ 84. The largest and most widely spread of the tribes of the Central
Belt is the Gond, a title which like that of K6I, has been extended to a
number of almost distinct communities. Some authorities trace the name
to Konda, the Telugu for hill, as in the case of the Kond or Kand tribe,
and they certainly cover the hill-country from Orissa westwards, with a
strong northern settlement in the Satpura and the south-west of the Cutia
Nagpur plateau. It has been already pointed out that their language ap-
proximates to the Kanarese rather than to the adjacent Telugu, but there
is little or no tradition of their earlier wanderings. The Raj-Gond, who
pushed up the Narbada and Kaimur, established a strong dominion on
the ruins of the Gauli dynasties, though it seems that they were in the

5- Ethnography.

neighbourhood long before that opportunity occurred, and were being
transformed into Nagbansi Rajputs even by the 4th century. The zenith
of their rule was from the i6th to the beginning of the 1 8th centuries,
when the Bhonslc overran their country and completely dispossessed them
of their power except in the hill fastnesses, which held out against all
comers. From the Kaimur the Gond passed eastwards into BaghL-lkhand
and the hills along the south of the Ganges valley. Here they are now
known as Majhvar or Manjhi, meaning headman, like Munda. In the Cutia
Nagpur States the Gond hold their land on military tenure, a fact which
seems to indicate that they were in possession before the present rulers.
All the northern and central Gond are more or less Brahmanised. The
upper classes, descendants of the former Chieftains, and the Chieftains
still holding petty States, claim to be Rajputs, and have for generations
intermarried with families of that order whose circumstances were in need
of reinforcement from some landed class better off than themselves. Under-
lying the prevailing beliefs, however, are the old tribal worship and customs,
and whilst Brahmans arc consulted as to lucky days and are brought in
to perform social ceremonies, the efficacy of the local priest and exorciser,
Pathari, Pradhan or Ojha, in practical dealings with the supernatural,
is everywhere acknowledged. In the south-east of the Gond country, from
Chattisgarh to Orissa, the tribes are far less Brahmanised, and live more
in the forest. The Maria form the principal section, and are found chiefly
in the Bastar State and the district of Canda. The majority of the Maria
are probably the wildest of the Gond, but on the outskirts of the hills
they are beginning, it is said to drop their designation for that of Koitar,
a more advanced section, and leading up to the title of Gond, without
any affix. The Koyi are less civilised than the Koitar, but the Bhatra,
or Bottada, to the east of the Gond tract, are nearly all Brahmanised,
some wearing the sacred thread, like the Raj section of the Gond. The
Halaba, originally from the Bastar State, have settled to a considerable
extent in the plain of Chattisgarh, and the further they get from the jungle
the more strenuously they disown connection with the Gond, and claim
to be an independent Brahmanic caste. As their main occupation is the
distillation of spirit from forest produce their claim is not encouraged by
the higher grades of the community to which they affiliate themselves. It
is not possible to give the numerical strength of all these sections of the
great Gond tribe or race, as at the Census the use of the general title
was very extensive. In 1S91 some detail was given, but on that occasion
also the value of the figures is diminished by the large number of un-
specified entries.

§ 85. Of the Dravidian tribes, next to the Gond come the Kand, or
Kond, with their kindred. The main body calls itself Kuyi, but the deri-
vation of both this and the ordinary title is uncertain. The Kand have
attracted a good deal of literary notice, partly due to their former practice
of human sacrifices and supposed advanced religious views. But the com-
munity is much subdivided and by no means uniform in its structure or
habits. There is, for instance, the usual division into the hill section, which
is untouched by Brahmanism, and that of the plains, which is adopting
both the language and religion of the Oriyas and Telugu respectively.
The Kand resemble the Gond in having pushed up northwards from the
southern outskirts of the ranges forming the abutment of the Central Belt
to the south-east. A further point of resemblance is the adoption of the

i

Castes and Caste-Groups. F. Hill Tribes.

name of the dominant tribe by bodies of artisans and menials who minister
to the former, so that, as in the case of the Nayar but on a smaller scale,
there are Gone! blacksmiths, drummers and cowherds, and Kand blacksmiths
and potters. The tribe lives by agriculture of the usual rude kind, but all
the Kand are also keen hunters, and very expert against game with their
bow and hatchet. They are very tenacious of their tribal rights over the
land they have once cleared, and in some cases, the whole of the village
land is held in common. The Kondu-Dora, on the contrary, who are
probably the southern branch of the same tribe, have lost hold of their
hills and are no more than a Brahmanic caste, speaking a mixture of
their old language and Telugu, and conforming to the ordinary local
customs. The Poroja, the meaning of whose title is uncertain, apparently
belong to the same stock as the Kand, but their language is held by
Dr. Grierson to be Gond, at least where the two communities live along-
side of each other. Elsewhere it is treated as a mixture of Kand and
Oriya. The tribe therefore, may be placed midway between the Gond and
the Kand. The Gadaba, again, are considered locally to be a branch of
the Poroja, and their subdivisions confirm this view. They are said to
have separate dialects of Oriya, all mutually unintelligible to the rest. In
the Linguistic Survey, however, the Gadaba language is classed with the
Savara, as southern Kol-Khervari. The tribe has no tradition of migration,
and lives by cultivation, one section working as carriers and labourers.
Their headmen act as their priests, and bear the same title as among the
Kand. The Jatapu are said to be Kand who have become in most re-
spects Brahmanised. Those residing in the hills speak Kand, but those
on the plain have taken to Telugu. The Jatapu, whilst observing the
orthodox rules as to marriage and diet, have never given up the old tribal
gods, to whom they sacrifice animals through their own priests, and keep
to their totemistic exogamous clans.

?? S6. There remains the Savara tribe, of which the greater portion
is now found in the Orissa hills and the adjacent wild country, under the
Central Provinces and iMadras. It has been already pointed out that as-
suming this tribe to represent the ancient Suari or Sabarae, they once
possessed a considerable dominion in the south Ganges valley. It is curious
to find even in the present day small communities bearing this name in
the very north of the Central Provinces and Bundelkhand, with no tradi-
tions of migration or former supremacy. The alternative designation of
the ISIale of Rajmahal, Sauria, has also been ascribed to some connection
with the Savara. Be this as it may, the detached body of the north-west
has lost all trace of its primitive religion and language, and is simply a
low caste of the ordinarj' Brahmanic type. Similarly, an offshoot of the main
Savara body which has settled in western Bengal, is gradually detaching
itself from the hill-dwellers of the tribe and employing Brahmans. It is
worthy of note that whereas the Savara in their native haunts seem to
be without exogamous subdivisions, those who have left the hills establish
them upon both totemistic and Brahmanic lines, borrowing the former,
probably, from some neighbouring tribe which preceded the Savara in
the valley. The wilder Savara have functional classes, such as the agri-
cultural, the metal-working, the weaving and the cane-working, but in-
tormation is not yet available as to the social distinctions implied in this
distribution. The Savara of the southern outskirts seem to be inclined
to branch off from their hill-comrades as they have done on the Bengal

5. Ethnocraphy.

side of the hills, and to gradually incorporate themselves with the Kapu,
or peasantry.

§ 87. (b) Western Belt (1,922,300). The Western branch of the
Kol tribes of the Central Belt differs considerably from those just reviewed,
owing, probably, to their having been driven into tracts which allow but
little room for cultivation, even on the methods adopted by the inhabitants
of the plateau. The link between the western tribes and the rest is found
in the Korva, a tribe Kol in its language, and by repute one of the
earliest settlers of the western parts of Ciitia Nagpur. The Korva, under its
western title of Kur or Korku, originated,' it is said, in the Mahadcv Hills,
and spread east and west. That they are amongst the oldest established
tribes seems certain, for other tribes get their priests from them in all
cases where village or local deities have to be appeased. The few of the
tribe who have risen to the rank of landed proprietors affect Brahmanism,
and set up as Rajj^uts, but the rest of the community, except, perhaps,
a few in the west, worship their ancestral ghosts and propitiate the ma-
lignant spirits of other people. In some of the States of Cutia Nagpur,
the Korva smelts iron and makes his own weapons and implements, but
this art is lost amongst those of the Satpura, who have to have recourse
to professionals for the large arrows which they use with considerable
skill at short ranges. Towards Betul and the Berar hills, the Korku are
divided into clans, the principal of which is called Muvasf. Further to the
west, this title is applied to the Bhils of the same range, and there is
doubtless some connection between the two. The Bhil has lost his tribal
language, and, except in the heart of the forest, much of his tribal religion.
Like the rest of his race, however, he maintains his respect for the old
pantheon as being more intimately and practically bound up with daily
life than the Puranic manifestations, even though the latter be brought
down to suit his requirements. The name of Bhil is generally derived from
a Dravidian word for bow, as in the case of the tree-tapping caste, Billava,
in Kanara. Probably this name, or at all events its interpretation, is modern
compared to the age of the community, but it certainly is applicable to
the Bhil of the present day, who in the forest, and even on the outskirts
thereof, is seldom without his weapons. In the west, the Bhil tribes are
divided, like the K61 of further east, into a Hill and a Plain section. The
latter, however, do not appear to aspire to more than a rudimentary form
of village settlement by themselves or than the duties of watchmen in
the larger villages of other castes. In the latter capacity, the permission
to retain his arms proved too strong a tem|)tation to be resisted when the
institution was first established, and the Bhil watchmen, with that marvellous
power of rapid concentration which distinguishes the tribe, were wont to
descend in force upon one of the villages exempt from their services. This
phase soon passed, and the Bhil is now a recognised part of the establish-
ment in the eastern villages of Gujarat. The Bhil worships the wood-spirits,
and in the west, at least, erects posts to them in the jungle, sacrificing
fowls and other offerings through a priest, generally of the tribe, whose
duty on other occasions is to discover the witches who seem to be pe-
culiarly active in this community. Some of the eastern Bhils have been
converted to Islam, especially those of the Tadvi clan, but their observance
of its tenets are very half-hearted, and the women especially, keep to
their former practices. In former days the Bhils held a good part of the
country north of their present hills, and were driven out by the Rajputs

i

Castes and Caste-Groups. F. Hill Tribes. 123

under pressure of the Muslim. Even now, they receive the respect due
to their former repute, an instance of which was given above in con-
nection with the enthronement of a Rajput Chief. There is strong reason
to think that the tribe was reinforced by the incorporation of refugee
Rajputs, who have left their marlc upon certain clans of Bhil, especially
in the south of Rajputana. In fact, the connection between the two is
said to have resulted in the formation of the Bhilala, now a separate
tribe. In regard to the relationship of the Bhil to the Korku, it may be
noted that the name of Mcvas, which is given to some of the Bhil tracts
in the west, is taken from the title of Muvasi, or Mavaca, by which they
are called, and which, as observed above, is the name of the western
branch of the Korku. Akin to the Bhil are the Dhanka, a tribe of south
Rajputana and Central India, the Patelia in the same region, and probably
of mixed origin, and the Tadvi and Pavada, which are both Bhil clans
locally separated from the main body, and settled in the Khandesh Satpura.
The Gamta, or Gamit, which nearly touch them on the north-west, seem
to be merely a superior class of Bhil, and not a separate tribe. The great
Koli tribe, which has been classed with the cultivators, contains, also,
more than one subdivision which still live in or near the forest, and have
not taken, like the rest, to either agriculture or seafaring pursuits. The
Naikada is probably one of these, as it is distinct from the Bhil, though
sharing the tastes and mode of life of the latter. The Naikada are found
along the south-west border of Rajputana and Central India, with colonies
in the forests of east Gujarat. They are by repute even worse neighbours
than the Bhil, and on several occasions have only been kept down by
force. For many years, however, they have been at peace, though showing
no disposition to abandon their primitive cultivation and their dependence
upon the jungle for their livelihood. The only advance they have made
is to engage under the Forest officials to cut and transport timber, instead
of working the jungle on their own wasteful plan. They pay homage to
Mata and Hanuman, as representing their own worship of nature and the
forest, but not only repudiate the services of the Brahman, but look upon
the murder of one of that order as an act of merit, and have the grim
saying, "By killing a caste-mark wearer, you feed a hundred." There is
a small tribe of much the same name, but settled at some distance from
the Naikada, called Nayak, which is unconnected with the Koli, and
seems to be the elder branch of the Dhundia caste, mentioned in con-
nection with agricultural labour. They are only found in the south-east
of Gujarat, where they live on the skirts of the forest, but not in it. The
Dhundia of the open country pay them respect at all formal ceremonies,
but do not eat or intermarry with them. The Nayak, moreover, have kept
up a good many of the tribal customs which the others have sloughed
off. Inter alia, they are terribly skilful and persistent on the local drum,
an accomplishment much appreciated at weddings and other festivals. The
Chodra of a little further north, are in appearance and customs much the
same as the Dhundia, but they have the tradition of having immigrated
to their present home from the south of Rajputana, whence they were
expelled along with some Rajput clans, by the Muslim. They resemble
the Dhundia in having taken to regular cultivation, with the addition of
cutting firewood from the forest for sale in the open country. Beyond their
worship of the village boundary-gods and their avoidance of Brahmans,
there is little to distinguish them from a low caste of Brahmanic cultivators.

124 5- Ethnography.

and they arc said to be gradually rising in position through their industry
and peaceful habits.

i} 88. (c) Sahyadri (367,600). The three or four small tribes of the
northern Sahyadri are almost contiguous to those just mentioned and
possibly are connected with some of them, though they have no traditions
as to their origin. The lowest of them, the Katkari or Kathodi, which
derives its name from the catechu it extracts in the forests, says it came
from the north, by which it means the forests of south Gujarat. The tribe
resembles the lower class of Bhil in appearance, but lacks the indepen-
dence and joviality of the predatory communities. The Katkari stick close
to the forests, and though they cultivate on a rude system, they never
take up land on a permanent tenure. They have their own gods and forms
of sacrifice, without reference to Brahmans. The tiger is an object of
special regard, as in Cutia Nagpur. Other tribes steer clear of the Katkarf,
not only because the latter are foul-feeders and remarkably dirty, but also
because of their reputation as sorcerers. It is worth noting that whilst
the principal demon of the locality is worshipped by the other tribes it
is rejjuted to be controlled by the Katkari, a difference implying the older
settlement of the latter tribe. The Varli (uplander), so called from being
supposed to have come from the country above the Ghats, are now re-
sident along the coast, but still in the forest. They are superior in ap-
pearance to the Katkari, and arc not adverse to permanent cultivation,
generally as subtenants upon the half-share system. The bulk of the tribe
also deal in jungle produce. They share some of the gods and ceremonies
of the Katkari, with the addition of Vaghoba, a tiger god recognised
by the lower Brahmanists. The latter, in turn, do not consider them as
altogether impure and enter their houses, or, at least, those of the Varli
who breed cattle. The third tribe, the Thakur, called for distinction,
the Ghat-Thakur, stands still higher in society, though except in being
a little more cleanly, the members of it have a strong physical resemblance
to the darker tribes of the north and east. They hold the same tribal
beliefs, too, and worship the mountain and tiger gods, but in their do-
mestic rites they make use of the Dcs'asth, or local Rlaratha Brahman.
The Thakur are settled in their own villages and possess land and cattle,
some of their community being fairly well-to-do. None of these three
tribes strays beyond its native haunts.

§ 89. (d) Nilgiri etc. (226,300). The comparatively small tribes of
the Nilgiri and the vicinity consist of descendants of a fugitive branch
of the Kurumban race and of communities the origin of which is uncertain.
To the former belong the Kuruman of the western slopes, who are the
same, except in locality, as the Kadu-Kurubar mentioned under the head
of shepherds. The general conjecture is that after the downfall of the
Shepherd dynasties of the south-east, some of the race fled into the jungle,
where they have since remained. The Irula, who inhabit the broken
country to the east of the Nilgiri, are apjiarently also of the same stock
if not belonging to the Coromandel Cencu tribes. Like the Kurubar or
Kurumban, they are divided into the section of the plain and that of the
forest. The former are more or less Brahmanised, live in villages and
work on the land. The others have the name of Villiyan, evidently derived
from the bow, their weapon of choice. Both sections worship the Kannimar
at an ant-hill in the jungle, these goddesses being probably the earliest
of all the Dravidian pantheon. The Toda and Kota belong to the table-

Castes and Caste-groups. F. Hill Tribes. 125

land of the Nilgiri, on which they have been isolated from pre-historic
time. Both apparently belong to the same stock, but the K(')ta admit their
inferiority to the others, though having turned out more adaptable to new
circumstances they appear to be more prosperous. The Toda arc essentially
a pastoral community, their sole wealth consisting of their stock of buffaloes.
Owing to their residing within an easy morning's walk of a popular hill-
station, also the seat of Government for the greater part of the year, the
tribe has received abundant notice, and has been to some extent cherished
as a valuable asset, being a specimen of what may be called "stall-fed
aborigines". There is some justification for this interest in the striking
difference in physical appearance between the Toda and most of the
surrounding population, as well as in their picturesque houses and mode
of life. It is probably, however, that they come from no great distance
from their present seat, and their language has been described as "old
Kanarese spoken in a gale", but it seems to have closer affinity to Tamil,
whilst the invocations more resemble Malayalam, with the Sanskritic strain
omitted. The Kota speak a different dialect, but the two tribes understand
each other. It is not improbable, therefore, that they both moved up to
the seclusion of the table-land from the Malabar forests in the neighbour-
hood of the Wainad or possibly even frt)m Coorg. In the ranges south of
the Nilgiri are found several small forest tribes, most of whom live in as
wild a state as the present conditions allow. The Kanikkar of Travan-
core are thought to be, like the Kurumban, the descendants of a race
once holding dominion over the surrounding plains, but driven to the hills
by invaders from the north. The title appears to indicate, like Bhumia
and its synonyms, the first claim to the soil, and this seems to be in
harmony with their position in relation to the Brahmanic castes below the
hills, who treat them as considerably purer than the menials of the village
or farm. They live by rude cultivation on the wood-ash system for a part
of the year, and then trust to hunting and the sale of jungle produce for
the rest. They are skilled in archery, and face elephants and tigers with
success. The iNIalayarayan, or Arayan of the hills, are more settled than
the Kanikkar, and have well-built villages, with considerable areas of
cultivated land. In some respects they bear a striking resemblance to the
Toda, as in not labouring for hire, but their reputation for practical sorcery
deprives them of the sympathy of the residents of the coast. Other hill-
tribes with the same title as the above or one closely resembling it, live
in the forests east of the Malabar district, with a similar fame as wizards
and casters of spells. All these tribes have been the subject of inquiries
in the course of the Ethnographic Survey, and till recently but little was
known about them.

ij 90. In the low ranges along the Coromandel coast, known as the
Eastern Ghats, a few wandering tribes are still to be found subsisting by
hunting, the collection of fruit and the sale of firewood to the villages
round. The Yanadi and the Cencu are connected with each other, and
according to the tradition among the former, the Cencu took refuge
amongst the Yanadi when driven from their home in the west. The Yanadi
call themselves Anadulu, or autochthonous. The two have the same tribal
deity, named Cencu, and worshij) without Brahmans or apparently priests
of any sort. It may be noted, also, that Cencu is the title of a subdivision
of the Gadaba tribe, further north, as well as of a section of the Yanadi,
and that the same name is given to the Irula in the uplands of Mysore.

126 5- Ethnography.

It is not improbable, therefore, that the tribes may be connected, and
that all came from the north, the Jrula having settled in the forests of the
transverse range uniting the eastern Ghats with the western, at the Nilgiri.
Another hypothesis is that the Yanadi may have been influenced in their
religion by the immigrant Cencu ; but the ethnology of all these tribes
rests largely on vague surmise. It used to be held that the languages
spoken by the Yanadi and Cencu were separate dialects of Telugu, but
it appears from recent inquiry that they are nothing more than the rural
vernacular spoken with a peculiar drawl and some differences in pro-
nunciation. Both tribes by preference live by what they can pick up in
the jungle, and sell fruit, honey and firewood in the villages of the plain.
The Cencu, too, occasionally breed cattle, and the Yanadi tell fortunes.
Both consider themselves above the leather-workers and lower menials
of the villages.

§ 91. Assam Tribes. The racial movements which have taken place
in this part of India were cursorily set forth in the Introduction. Owing,
no doubt, to the comparatively recent date at which successive settlements
have occurred, and also, to the natural isolation of some of the tracts,
which have thus been unaffected by alien inroads, the racial concentration
coincides, as a rule, with the geographical position. There are exceptions,
of course, as in the Central Belt, where a tribe has been cut off from
its fellows, or the new-comers have been unable to effect a continuous
occupation, but in most cases the tribes in question can be dealt with in
groups which are geographical as well as racial.

The general results of the Ethnographical Survey of Assam have
not yet been published (1909), but several valuable monographs upon parti-
cular tribes have been prepared by local officers specially qualified for
the task, and some of these have been utilised in the last three Census
Reports. The numerical strength of the tribes, however, which it is the
main object of the Census to discover, is not altogether satisfactorily re-
presented by the returns, partly because of the variety of language current
amongst these communities, which has the result of giving to many of
the latter a title unknown within their own body. The influence of
Brahmanism, moreover, upon the numerous less civilised tribes by which
it is here surrounded, turns the scale adversely to accurate ethnographic
nomenclature. Members of a tribe who decide upon conformity with
Brahmanic observances are apt to signify their breach with the past by
adopting the name of an existing caste, with or without a qualifying
epithet. Taking an example from one of the larger communities, a Kacari
does not make use of that name, but calls himself Bara, and when he is
dallying with the outworks of Brahmanism, he is a Saraniya, or a Saraniya
Koc. Once the plunge taken, the prefix is dropped, and he becomes
Koc. In due course, if he thrives, he dies Rajbansi. As the same course
is followed by the Lalung, Mikir and Garo tribes, the identity of the
convert is lost in an all-embracing title, once racial, but now sunk into
nothing more than the designation of a loosely knit and heterogeneous
Brahmanic caste. Thus the remarkable variation in the numbers returned
for a tribe between one Census and another is attributable to little more
than additional care in the discrimination between local terms, and, on
the whole, the later enumeration may fairly be taken as more correct
than its predecessor. There are other causes of variation, but they are
exceptional. One tribe suffered more severely than others from the serious

Castes and Caste-Groups. F. Hill Tribes. 127

epidemic, called the "black disease", which ravaged the valley a few
years back : another, the bulk of which resides beyond the frontier, may
have sent more or fewer immigrants into British territory.- elsewhere, the
Census was extended to tracts in which it was not possible to conduct
the operations ten years before, and so on. Even now, there are tribes
of considerable importance dwelling in the north-eastern and eastern hills,
which have not yet been enumerated.

The information available, then, extends to the main Bodo group of
the Brahmaputra valley and the Garo Hills; the Khasi of the hills bearing
that name; the Mikir, similarly identifiable to the east, the Naga and the
Kuki-Lushei, to the east and south, and to the small San tribes in the
north-east. It is imperfect in the case of the Naga and the Cin, and also
as regards the Himalayan tribes skirting the northern edge of the Brahma-
putra valley. Of all the tribes comprised in these groups not more than"
two or three claim to have been always where they now are, and even
in these cases it is probable that it is only the tradition of immigration
from the north-east which has been lost. The different waves of migration
which landed most of them in their present home took place at such long
intervals and from such various sources that there are few general cha-
racteristics common to the Mongoloidic population in the aggregate. In
regard to religion, most of them profess the belief in one deit^- above
the rest, but as he is passively benevolent only, the tribal worship has
to be directed chiefly to the propitiation of local agencies which are
actively malignant. This object is attained by the sacrifice of some animal,
varying according to the occasion from a fowl to a buffalo, with a pig
as a good working intermediate offering. The tribes of the valley have
in some cases a levitical clan of priests, but generally, the officiator at
the ceremonies is a medicine-man, either elected or hereditary, belonging
to the tribe or clan. Occasionally, especially in the eastern hills, the
village headman presides. In many tribes there is a belief in a future
state, mixed with the possibility of the return of ghosts of deceased
members of the tribe. Those who have seen a good deal of the every-
day life of these bodies testify to their sound notions of tribal honour
and morality, though in regard to strangers their institutions are apt to
prove repellent. Amongst all the Naga tribes, for instance, and some of
the Kuki and Cin, the custom of collecting the heads of members of other
communities is only kept down where the British Government has established
itself firmly, the inclination towards this form of vanity being as strong
as ever. Other tribes used habitually to raid their neighbours for girls
and boys to be kept as household slaves, the offspring being formed
into a separate community, as is the case in the west of India. The
village and its constitution, too, presents many interesting points of diffe-
rence amongst the wilder tribes, and whilst most of the latter are content
with the rude jungle cultivation which prevails amongst the Kol tribes,
others have struck out a line of their own, and grow superior crops, in
one case by means of an elaborate and almost unique system of irrigation.
Some tribes are divided into exogamous clans, mostly totemistic, so far
as is known at present; others live in village communities, each under
its own ruler, independent of the rest. These, it may be assumed, are
closely stockaded and in a good situation for defence. Others acknowledge
the sway of a local Chieftain owning several such villages. The unrege-
nerate tribesman of the valley, builds his house on a platform and enters

128 5- Ethnography.

it by a ladder; whilst on conversion, he builds on ground-level and goes
in by a door. Omens, divination and witchcraft prevail throughout.

§ 92. (a) Bodo (817,300). Dealing first with the Brahmaputra valley,
the principal tribe still in occupation is the Bodo, or Kacari. It is now
chiefly found along the northern bank, from the western limit of the
Province to the Darrang district. Formerly, however, it possessed territory
far to the east and south, and in the latter direction it is still the prin-
cipal population of the Hill Kacar tract, received, it is said, as a dowry
from Tipparah, in the palmy days of Bodo dominion. The Bodo are
undoubtedly of trans-Himalayan origin, but it is uncertain by what route
and stages they reached the valley. It is said that they first rose into
power in the north-east of the latter tract, and spread down the river
and across it as they approached the plains. They have no traditions,
and belong to the peoples of whom it has been said -their languages
are their history*. Upon that basis, they are allied to the Garo, Mcc,
Rabha, Lalung and Tipparah tribes, and also to the K6c. In the present
day the Bodo are a sturdy, independent, and remarkably clannish com-
munity of labourers. They have none of the objections of the hill tribes
to seasonal migration, and frequent in large numbers the teagardens of
the upper valley. Their tribal subdivision seems to be different in the
Hill country from what it is in the valley. In the former exogamous
sections are strictly maintained, but in the latter, such as there are seem
to be weakening in vigour, and though nominally kept up, and the clan
name still descending in the male line, marriages are no longer regulated
in accordance with them, nor is the totemistic prohibition regarded,
except, perhaps, to the extent that the tiger clan are not allowed to
abuse that animal when shot, as the rest do. The number of the tribal
population is considerably more than the figure here quoted, since many
of the converts to Brahmanism, as above stated, do not retain their tribal
name, and whole villages in Upper Assam are inhabited by pure Bodo,
though that title is not returned by a single family. Across the Brahma-
putra, mainly in the range bearing their name, are the Garo. These claim
to be autochthonous, but their tongue and customs indicate a close rela-
tionship to the Bodo and to the Lalung, a neighbouring tribe on the east,
of the same race. The Garo are not found far from their hills, but a few
thousands have made their way into the adjacent district of Bengal and
across the river into Goalpara. The tribe is much subdivided. There are
four main clans, each of which has its numerous exogamous sections. In
religion the Garo resemble the Bodo, and have the same system of pro-
pitiating the malignant deities through the Kamal, a non-hereditary priest,
corresponding to the Deori of the others. The Lalung are now found
on the north slope of the Jaintya hills, spreading into the valley bordering
the Mikir country, with apparently a tendency to advance still more to
the eastwards. Traditions differ as to their original home. Some clans
say they came from the south bank of the Brahmaputra, others that they
are wholly Jaintya, and have never lived anywhere else. They do not
appear to have been in the low country when the Ahom invasion took
place, in the 13 th and 14th centuries. It is said that they are succumbing
to the influence of Brahmanism, but if this be so, they must cither change
their name on conversion or the enumerators at the Census must ignore
their tendencies, as they are recorded as wholly Animistic in their beliefs.
There is no doubt, however, that they are dropping their tribal language

I

I

Castes and Caste-Groups. F. Hill Tribes. 129

in favour of that of the lowlands. The number of exogamous subdivisions
into which the tribe is split up is ver>' large, and it does not appear that
they are usually totemistic as a rule, but are named after some peculiarity of
the founder. The Rabha is a tribe certainly of Bodu blood but whether a
distinct community, allied to the Garo, or merely a branch of the Bodo,
alongside of whom it is chiefly found, is not determined. Some have
thought that the Rabha was a name given to a half-converted Garo or
Kacari, and it is certain that there are Garo who have become Rabha
without passing into Brahmanists, just as the Kacari passes into the same
community without proceeding to the grade of Koc. The converts constitute
a sub-tribe by themselves. On the whole, the Rabha hold themselves to
be above the Bodo, but marry girls from the latter. The Bodo, on the
other hand, does not marry a Rabha without some purificatory rites. The
special dialect of the Rabha is said to be dying out in favour of Assamese,
and the people who join the Brahmanists call themselves Koc, so the
tribe is on the way to extinction. The Mec live mostly in the Tarai on
the west of the Brahmaputra, partly in Assam, partly in Bengal. From
their comparatively fair complexion and Mongoloidic features they are
affiliated to the Bodo, though they have no tradition of ever having lived
out of the Tarai. They intermarried with the Koc Chiefs, a fact which
seems to support the theory of Bodo relationship. Towards the west, in
Bengal, they are chiefly Brahmanists, and divided into two endogamous
sub-tribes, one of which intermarries with the Dhimal, a tribe of different
race, possibly Kul or sub-Himalayan Nepali. The Assam Mec have kept
up customs much resembling those of the Lalung. A small tribe, akin to
the Garo and Bodo, called Hajong, inhabits the southern slopes of the
Garo hills, and has made its way into the Surma valley. This descent
into the plain appears to have resulted in the formation of two clans, the
upper, which remains true to its tribal ways of life, and the Brahmanised
community of the valley. The latter have also abandoned their tribal
dialect in favour of a corrupt form of Bengali, the others speaking one
of the varieties of Garo. Detached from the main body of the Bodo is
the Mriing, called Tipparah by the Bengali, and now inhabiting the hills
near the little State called by the latter name. A few of them are found
in the Surma valley, but most of these are said to be immigrants of quite
recent arrival. Formerly the connection between the tribes was closer,
as the Chiefs of Kacar and Tipparah intermarried. Now, the only link
is that of language, as the bulk of the JMrung are Brahmanised, the Chief
claiming to be a Rajput, and the nobles to belong to the Rajbansi order.
The tribe is much subdivided, some clans holding an position far above
that of the labourers and rude cultivators of the interior. Many of them
are much fairer than any of their neighbours, and this, with their Mon-
goloidic features and Bodo speech, seems to connect them with the
Brahmaputra rather than with the hills of Arakan. Last of the tribes
coming within this group is the formerly dominant community of the
Ciitiya, which, however, repudiates the connection with the Bodo indicated
by their language. They are said in the ancient Assam histories to have
come down from the north-east, and to have founded a kingdom in that
corner of the valley afterwards expanding southwards into Sibsagar and
Nowgong. They came into contact with the Ahom, and were dethroned
in 1500. Before that date they were in part Brahmanised, and their com-
munity is now divided into the Brahmanic, the Ahom, the Borahi, or pork-

Indo-Aryan Research. II. 5. 9

130 5- Ethnography.

eaters, and the Dcori, or Levitical body. The two first have been for
some time almost completely converted to Brahmanism, and the fourth,
though standing out for some generations, has now succumbed, on social
considerations, it is said, rather than by religious conviction. The Borahi
are a lower class and were the first to fall before the Ahom, who reduced
them to a servile condition. They are now apparently almost extinct as
a separate community. The Cutiya have lost, along with their religion,
their tribal language, which is closely allied to that of the Bodo. They
are no doubt of the same origin, but ha.ve long been separated politically
as well as geographically, and occupied in upper Assam the same domi-
nant position which the Bodo held lower down the river. At present the
majority of the Cutiya are found to the south of the Brahmaputra, in
Sibsagar, Nowgong, and Lakhimi)ur. The Deori have remained in and
about their original seat in the extreme north-east. The princii)al object
of their worship is Durga, who was enthroned in place of the numerous
evil spirits to whom the tribe paid homage before their conversion. Even
now, the services of Brahmans are not called for, and the sacrifices are
performed by the Deori and his assistants. The more Mongoloidic appea-
rance of the remnants of the Deori clan seem to indicate that they have
kept themselves freer from intercourse with the Bodo and Ahom than
the rest of the Cutiya. One of their social peculiarities worth mentioning
is the habit of lodging a whole family under one roof, enlarging the
building as the numbers increase, until sometimes more than a hundred
persons are thus sheltered. Their professed Brahmanism sits very lightly
upon both priest and layman, and is almost confined indeed to the obser-
vance of the initiatory injunction of offering prayers, keeping secret the
instructions of the Gosal and paying their annual fee to that functionary.
§ 93. (b) The Himalayan tribes (48,000). Though few of these, and
those not the more important, have descended into British territory, they
may be briefly mentioned here owing to some alleged connection between
them and the Bodo race, a tie, however, which has long been severed.
The Miri is the only tribe which has settled in British territory to any
considerable extent. It is found in the Sibsagar and Lakhimpur districts,
and seems to be receiving recruits from the hills to the north of the
latter and from Darrang. The Miri say that they were invited down by
the Ahom Chief at the end of the 1 8th century, in order to help him
against the invading Khamti, and settled on the outskirts of the Naga
hills, by the Disang. They have preserved their original type in spite of
considerable defections from the tribal religion. Brahmanism, however,
affects them but superficially, and those who have nominally accepted the
guidance of the Gosaf, are now, it is said reverting, because the change
of faith has not induced the settled population of the valley to intermarry
with them or to accord them any better position than before they paid
toll to their spiritual adviser and renounced beef. In any case they do
not entrust their principal sacrifices to other than their own tribal priests.
They are good cultivators, and their women folk work with them in the
field. The Hill Miri, who only visit the plains for the purpose of trading,
are much less advanced, and have a somewhat different worship and belief
from the others. All the Miri are connected with the Abor, a stronger
race, and it is conjectured that it was the pressure of these northern
kinsfolk which drove the Miri to the lowlands. It is advisable to note that
the name of Miri which means Middlemen in Assamese, is not known to

i

Castes and Caste-Groups. F. Hill Tribes. 131

the tribe itself, any more than that of Abor is recognised except in the
valley. The latter means Independent, and is thus appropriate enough.
Both tribes speak of themselves by their clan, without any more general
designation. The Abor have not yet settled to any great extent within
British territory, but have more than once made raids therein, which resulted
in punitive expeditions. Their clans are very numerous, but are remarkable
for the unanimity with which they combine into a tribal whole for purposes
of resistance or plunder. They used to be keen on the capture of girls
and boys, whom they kept as household slaves themselves, and sold for
the same purpose to their kinsmen, the Daphla, who live the other side
of the Miri, on the west. The Daphla, who call themselves Nyising — the
meaning of both terms being unknown — regard the Abor as the leading
tribe of their race and the ^liri as poor relations, and all three speak
much the same tongue, and to some extent, have the same titles for their
sub-tribes. The religions present the same general features, and the .\bor
and Daphla have not been reached even by the light touch of the Miri
form of Brahmanism. The Aka, a tribe adjacent to the Daphla on the
west, though mainly of the old faith, has a few members who are reported
to have been converted by one of their Chiefs, who chanced to be com-
pelled to serve a certain time in a British jail, where his convictions were
modified by a persuasive Gosai. The Aka, though generally thought to
belong to the Abor-Miri race, differ considerably from both of these in
appearance, and show but little tendency to settle in the lower ranges.
On the contrary, they are in close relations with the Tibet authorities on
the other side. They are a warlike community, and in addition to their
general title which is not used by them, and the meaning of which is un-
known, they have two subdivisions, each of which is known to the Assamese
by a title implying plunder.

§ 94. (c) The Khasi and Sainteng (159,500). These tribes belong to
the same stock and speak the same language. The former reside in the
western portion of the range bearing their name, whilst the Sainteng share
with the Lalung the Jaintya portion of the same range. In treating of
languages it was pointed out that these two, with two smaller communities
of the same tract, appear to be the remnants of a wave of the Mon-speaking
race, left stranded by the main body. They have no traditions of any other
home, and differ considerably from the surrounding tribes in customs as
in speech. The numerous exogamous Khasi clans, for instance, are based
upon descent from a female ancestor. Inheritance is in the female line,
and the woman is the head of the family. No money or gift passes on
marriage, and the young couple do not set up house until a child is born.
The religion is the usual propitiation of evil spirits, with a faint and dim
notion of a future state in which husband and wife rejoin each other,
unless a widow has married again, in which case she belongs to her second.
Of late the Khasi have been converted in considerable numbers to Christia-
nity, and a few have become Brahmanists. The Sainteng show less dis-
position to change. On the other hand, though sharing the religion and
customs of the Khasi, they appear to have received a greater admixture
of foreign blood, due, it is thought, to the greater accessibility of the
Jaintya hills from the plains on the south. The Khasi, again, are divided
into petty States or independent groups of villages, each forming a little
republic under its own head. In the sister hills, the country' is altogether
under the Chief of Jaintya, who appoints twelve local officials to carry on

132 5- Ethnography.

the village affairs. The Chief himself is a Brahmanist, but his example,
as just mentioned, has not been contagious, and the annual tribal devil-
drive, in which every male takes part, is as popular as ever.

(5 95. (d) The Mikir (87,300). This tribe inhabits the lower portion
of the Khasi range on the north-east and has spread over the plain to
the east, up to the Naga hills. The traditions it has regarding its former
home are vague and valueless, but it pmbably occupied the low range
which goes by its name after leaving the Jaintya hills. From the language,
it is supposed to have some affinity to the Naga race, though in habits
and appearance it might well be affiliated to the Bodo. The Mikir call
themselves Arleng, meaning simply Man, an appellation so common amongst
forest tribes that it affords no guide to identification. They are subdivided
into several large sections which may, but do not, intermarry. Their chief
god is benevolent and powerful, but his subordinates, though theoretically
less in authority, arc more active, and generally work mischief. The sacri-
fices to them, accordingly, are more frequent. They are conducted by
priests who are selected from the elders of the clan, whether men or
women. The Mikir are excellent agriculturists in their own line and keen
traders in disposing of their crops. They are peculiar amongst their kind
in these parts in not congregating in large villages, but in building a few
large houses close to their fields. They are great breeders of buffaloes,
but, like almost all hill-tribes, K61 or Mongoloid, they abstain from making
use of milk. Until recently they had resisted the temptation to embrace
Brahmanism, but of late a certain number on the southern limits of their
tract have begun, it is said, to observe certain restrictions in diet when
out of their own village. Physically, the Mikir stand second to the Bodo
and above the rest of the tribes here mentioned. Whatever may be their
connection with the Naga or other races, they themselves deny any rela-
tionship with their neighbours.

§ 96. (e) The Naga tribes (i62,Sooj. This name is applied by the
outside world of Assam to a collection of tribes occupying a considerable
hilly region between Manipur and the south bank of the Brahmaputra.
The communities themselves know of no general title, and their tribal
designations are seldom those by which they are called by their neighbours.
A large amount of information about them has been collected in connection
with the Ethnographic Survey, and until this is given to the world, no
adequate account of them is available. It is probable that they reached
their present locality from two directions. One branch came down from
the north-east, whilst a later section doubled back northwards, after having
spent some time alongside of the Kuki and other tribes, to the south. The
largest tribe, as far as is at present known, is the Angami, called Tengima
by its own members. It is settled along the western ranges of the hills,
and is one of the communities said to have come from the south. The
Tengima reside in unusually large villages, some containing as many as
800 houses. The villages are set upon a hill, and carefully stockaded and
guarded against attack. The unit of the tribe is not, however, the village,
but a subdivision of the population thus concentrated, called Khel or Tcpfu,
exogamous, and said to be derived from a single ancestor. Faction-fights
between these bodies are frequent and used to be bloody, as outside aid
was called in to take part. The large size of their villages is probably the
result of their adoption, apparently from the Manipuri, of the system of
permanent cultivation by irrigated channels, carried with extraordinary

I

Castes and Caste-Groups. F. Hill Tribes. 133

skill and labour round the slojics of the hills. They have the usual vague
tribal belief in a supreme god and a future state, though they have not
formulated their notions of what happens to the soul when it leaves the
body. Their worship is devoted to the propitiation of the spirits of nature,
who inhabit pools, trees and rocks, and cause illnesses. The beginning
and the end of harvest are celebrated, as in the valley, with elaborate
festivals. The Ao Naga tribe came from the north, and is settled to the
north-east of the hills. The men are inferior to the Tcngima in physique
and in their way of life, but their buildings and villages arc, if anything,
superior. Beyond a few special tribal customs, the two tribes have much
the same beliefs and practices. The Ao arc really two communities, the
Cungli and Mongsen, which speak different dialects and intermarry, each
having its own e.xogamous sub-sections. The enslavement of members of
neighbouring tribes used to be a regular custom, now, of course more
or less suppressed. The victims were treated well, except when paid over
as fine or ransom to another village, when they were usually sacrificed.
The villages, though nominally governed by a headman, are in practice
independent democratic units. The Sema, or Sima, village, on the con-
trary, under the adjacent tribe, has a hereditary headman, or Chief, endowed
with considerable authority and privileges. This tribe came from the south
east, near Kohima, and has occupied a considerable tract round its present
settlement. The Sima are more akin to the Tengima than to any other
of the local tribes, but are distinguished even among the Naga, for their
barbarism and ferocity. They used to prey upon the lands of the Ao,
but having been headed off under British control, they are spreading
eastwards, over a wilder country'. The Lhota, in contradistinction to the
Sima, are a quiet and industrious people, though they adhere to the old
method of cultivating on burnt patches of jungle. They manage, neverthe-
less, to grow a good deal of cotton, which they convey themselves to the
river for sale. In habits they resemble the Rengma, their neighbours.
A section of the latter, being evilly entreated by other tribes, sought the
lower hills, east of the Mikir, where they alone of all the Naga have taken
to something approaching the life of the population of the plains. As to
the large number of tribes in this group which live in the interior and
south of the hills, little information beyond their titles is at present available.
§ 97. i^f) The Kuki tribes (200,200). Almost the same remark applies
to these, with the exception of the Manipuri and Lu.sei. In tl'.c Kacar
hills are found some called the "Old Kuki" (67,200), who were driven
north by others of the same race, who, in turn, were being pressed hardly
by the Lusci. The principal tribes of the former are the Rangkol and
the Bete. They are subdivided into eight social grades, like castes, with
the all-important difference that they intermarry with each other and with
other tribes. The existence of exogamous clans is probable, but the
nomenclature obtained at the Census throws no light upon this point. The
Rangkol, and probably the other tribes, worship one chief and several
minor deities, and select one t)f their own clan to serve as priest. In Kacar
they are beginning to mould their diet upon Brahmanic lines but not so
as to interfere materially with their ancestral habits. They differ from the
other Kuki in having no Chief, but they elect a headman for each village
to manage its affairs. The population of Manipur is divided into four
tribes, the Khumal, the Luyang, the Ningthauja or Meithei, and the
Mayarang, of which the Meithei (69,400) seems to have absorbed the others.

134 5- Ethnography.

and is used as a general title by the inhabitants. The exogamous sub-
divisions of the tribes, however, are still in existence, and seem to consist
of the descendants of an individual, by whose trade or nickname the section
is called. In 1720, the then Chief, called by the Muslim title of Gharib
Navaz, was persuaded by some Brahmans at his court that he and his
subjects were K.satriya of the Lunar race. The monarch thereui)on embraced
their creed and was invested with the sacred thread, and with him a large
number of his people. Since then, not only have most of the Meithci
become K.satriya, but the rank has been conferred by the Chief upon a
plentiful supply of recruits from the surrounding Kuki and Naga tribes.
The result is that at the Census only 33 of the inhabitants of the State
returned the tribal name, whilst the 33,000 Manipuri found on the record
are bastard Bengali enumerated in Kacar and its vicinity. The Brahmans
wht) first entered the State upon their mission of conversion were given
wives of the class of Kei, or Naga slaves of the Chief, into which body
their descendants also married, so that the sacerdotal caste docs not bear
any special title to respect in the eyes of the local K.satriya, to whom many
of them act as cooks, for the convert is most particular as to diet and
intercourse with his inferiors. Nevertheless, they have 300 deities of the
old worship who are still propitiated through the native priest, or Maiba,
and in every house hangs the basket containing the household god. The
connection of the ruling family with the Jadav clan has naturally attracted
the Manipuri K.satriya to Mathura, the centre of Krsna-worship, where a
small colony of them appears to reside. They also observe the great Kr.sna
festivals in their native country. The Loi clan of the papulation seem to
be descended from the Mayarang, and now to constitute a sort of receptacle
for anyone degraded from the K.satriya class. The Loi are the helots and
labourers of the State, and the original families of the clan have their
own dialect. It seems, however, that a Loi who embraces Brahmanism
and has never been degraded from any other position, may be made at
once a Ksatriya.

§ 98. (gj The Lu§ei (63,600). This people, who call themselves Dulien,
are of the same race as the Thado, or Kuki, whom they drove out some sixty
or seventy years ago. Long previous to that date, however, a Chief of
the Lu.sei had subjugated most of the hill villages around him, and his
descendants are said to be the progenitors of the present numerous Chief-
tains who rule the tract. The clans and subdivisions are many, but they
seem constantly to be being absorbed or reformed, always with reference
to connection with the eponymous founder. Each village is under one of
these petty Chieftains, who is entirely independent but has recognised
duties towards his fellow villagers, and in return receives a certain share
of each man's rice crop. The only remedy against a too despotic headman
is to flit, and transfer allegiance to another village. The village itself is
stockaded, like those of the Naga, but is laid out differently, the streets
radiating from a square in the centre, in front of the house of the Chieftain.
Except in detail, the religion of the Lusci does not materially diflfer from
that of the tribes just mentioned. Like most of the Kuki, the Lu.sei is
a keen and expert hunter and snarer, and seems to carry into his warfare
the qualities which makes him successful against wild animals, for he
rarely attacks except from ambush or by a surprise. The tribe is not given
to head-hunting for the mere sake of the trophy, but cuts off the head of
his enemy in order to prove to the women at home that he actually killed

Castes and Caste-Groups. F. Hill Tribes. 135

him. South of the Lusei Hills, the tribes almost entirely belong to Burmese
races, with which this review is not concerned.

§ 99. (li) The San tribes (4,600). The portion of this great race
which has found a home in British India is but small, and, with one
exception, of comparatively recent settlement. The break-up by the Burmese
of the Mau San dominion on the upper Irawadi, about 1760, obliged
several small bodies of different tribes to cross the Patkai, and settle east
of Sadiya, on the Brahmaputra. Amongst these are the Khamti, Turung,
Nora and Phakial. The Khamti were originally connected with the Ahom,
who will be mentioned later, and it was with the permission of the Ahom
Chief that the former obtained a foothold in Assam. They encroached,
however, got into trouble about their practice of raiding for slaves, and
were finally scattered about 70 years ago, many returning across the hills
to the Irawadi. A few years later another colony api>carcd and settled
in the same tract, where they now are. The Phakial also belonged to
the Mogaung kingdom, and had to leave when the Burmese overran their
country. They did not make direct for Assam, but halted on the way.
Being probably pressed by the Singpho, or Kacen, they accepted the
invitation of the Ahom to settle along the Dihing, and afterwards near
Jorhat, from which, however, they withdrew when the Burmese entered
Assam. The Nora belong to one of the tribes of the Ahom which elected
to remain on the east of the range when the main body crossed into Assam.
They are also called Khamjang, from one of their halting places in the
north-east. From this they were ejected about a century ago by the Singpho,
and came into Assam for safety. It is said by the Turung, another tribe
of the same origin, that the Nora, having settled in the valley, sent for
them to join the colony, and as they were oppressed by Kacen, they came.
On the way, however, they were taken prisoners and enslaved by the
Singpho, and were only released on the arrival of a British expedition
in 1825. They intermarried with their captors and are accordingly looked
down upon by the Nora, still more by the Khamti, who stand at the head
of the San community of Assam. Turung brides are taken by the others,
but none are given in return. All the above tribes are Buddhist and have
their own priests. The Alton, a small band of refugees from the San
court of Mungkong, settled in two bodies, one near the others of their
race, and the other in the Naga hills. Both, though professing Buddhism,
are gradually becoming Brahmanised, alike in creed and language. The
Census figures for these small communities are anything but accurate, as
many are set down simply as San, and others as Buddhist, without any
tribal title. Finally, there are the Ahom, the only tribe of long settlement
and political importance. They have been mentioned more than once in
connection with tritial religion and language, having abandoned their
tradition and practice in regard to both. They have preserved, however,
a very complete series of histories of their career. From these it appears
that they left Mogaung on the Irawadi about 1228, in consequence of a
dynastic dispute, and crossed the Patkai into the north-east corner of the
province which now bears their name. By 1500 they had subjugated the
Cutiya ; and forty years later, the Kacari or Bodo dominion fell to them.
They recovered from a severe defeat at the hands of the Koc, and repulsed
on several occasions an invasion by the Muslim, getting possession of the
valley as far west as Gauhati, and later, to near Goalpara. Their decline
set in on the conversion of the Chief to Brahmanism. Discontent arose

136 5- Ethnography.

amongst those who would not follow his example. Some rebelled ; the
seat of government was withdrawn down the valley; the Burmese were
called in, and ended in absorbing the whole kingdom, until the British
took possession. It seems that the Ahom were divided into classes but
whether these were endogamous or not is uncertain. The highest class
Comprises the Chiefs family and six or seven others of rank. The middle
class is divided functionally, and the third comprises all who are bound
to render services to the Chief. There were also Levitical or priestly
families. In the present day the distinctions based on occupation and on
service formerly rendered are dying out. The whole tribe has become
to a greater or less extent Brahmanised; that is, the spiritual authority of
a Gosai is acknowledged, and some changes in diet are gradually adopted.
The priests, as in the case of the Cutiya, stood out for some time longer
than the rest, but have now conformed. It is curious that whilst the little
that remains of the sacred writings of the Ahom is in a language closely
resembling that of the Khamti, the y\h6m were never Buddhists. It may
be inferred from this, perhaps, that the latter had not reached the upper
valley by the 13 th century. Nowadays, the Ahom are all nominally Brah-
manists except about 400, and it is said to be only a matter of time for
the whole tribe to be absorbed into the various castes of the valley.

§ 100. The Singpho (1,800). So few families of the great Kacen race
are found within the borders of India, as the limits of that country are
here understood, that the only reason for mentioning them is the reference
made above to their interception of bodies of immigrants on their way
to Assam. About a century ago a small colony of the northern Kaccn
made their way into the same corner of the valley as the rest of the
Irawadi races had done, and there they have remained, under their Assamese
designation of Singpho, or "the Men". The main feature of interest in
connection with them is that the offspring of their alien slaves, who form
a separate community called Doania, now outnumber their former lords
and masters. Both are Buddhist in the main, but the Doania are inclining
towards Brahmanism. About 340 are returned under their tribal religion.

§ loi. Himalayan (Nepali) tribes (218,600). Of the tribes coming
within this group only a few are settled in British territory, and the rest
belong to Nepal, where no Census has been taken. Almost all of the
former class are concentrated in Sikkim, Darjiling and the immediate
neighbourhood, whilst the Nepali subjects are either sojourners in or about
the same locality, or are serving in the Gorkha regiments in Assam. The
Lepca, or, as they call themselves, the Rong, claim to be the original
inhabitants of Sikkim, though one of their subdivisions is said to have come
down from the Chinese frontier. The Khambu and Li mbu assert them-
selves to belong to the Kirata race, a pretension which is not allowed
by the Yakha, who would limit the territory associated with that ancient
title to the tract between the Diid-Kosi and the Tambor river, where they
live themselves, along with a tribe known there as Jimdar, or Rais. This
title, however, has been appropriated by the Khambu living in the Darjiling
territory, but it would not he allowed to them across the Nepal frontier.
The Limbu touch the Kirata tract on the west, the Khambu on the north,
and the Lepca on the east. The Limbu are amongst the earliest inhabi-
tants of the country where they are still found, and from their appearance
it seems that they are originally from Tibet. Their petty Chieftains were
in power towards the end of the i8th century, when the Gorkha occupied

i

Castes and Caste-Groups. F. Hill Tribes. 137

Nepal, and incorporated the Kirata land with their new acquisition, after
a stout resistance from the Limbu. The latter take rank amongst the
Kirata tribes after the Khambu and before the Yakha, though, as above
remarked, in Nepal the order may be different as regards the Yakha.
A certain number of the Limbu have entered into close relations with the
Lepca, intermarrying with them and eating their food, a course which
amongst the other Kirata places them outside their fellows. At the same
time, it appears that the Lepca, Miirmi and other Himalayan Mongoloids
are admitted into the Limbu ranks after certain ceremonies, whilst the
Khambu and Yakha may be adopted without such formalities. The Limbu
have their own priests as well as using the exorcists, or Bijua, common
to all the tribes of the neighbourhood. They indifferently profess S'aivism
when amongst Brahmanic castes and employ the Lama at a higher altitude.
Probably their real creed is that of old Tibet. Their kinsfolk and neigh-
bours, the Khambu, live on the southern range of the Himalayan system,
where those who own land call themselves Jimdar, .so that this title has
been merged in the general tribal designation at the Census, without
reference to the claims of the Yakha mentioned above. They profess
Brahmanism, but employ no Brahmans, and serve an ancestral deity through
Home, priests corresponding to the Bijua of the other Tibetan communities.
They seem to have some faint reminiscence of Buddhism in portions of
their worship, and may once have passed through a phase of that creed,
like many of the Himalayan tribes. They intermarry with a beef-eating
tribe of Khambu from the north of the main range, and on that account,
irrespective of the quarrel about nomenclature, are kept at arm's length
by the Kirata of the west. These last, as well as the lower tribes of Kirata,
such as Hayu, Thami, and Danuar, of the Tarai, are only sparse and
occasional residents in British India. The Lepca probably represent two
different immigrations from Tibet or its eastern frontier, but the sections
are now amalgamated. Amongst the clans, however, two stand above the
rest, and do not intermarry with other Lepca or with Limbu, and it is
possible that these are the descendants of the semi-Chinese band introduced
along with one of the Sikkim Chiefs from across the Tsan-pu. In the
present day, the Lepca is working a little more steadily than he was
accustomed to do before the British occupied Darjiling, but he still objects
to remaining more than a few years in one locality, and after a season
or two of careless cultivation, moves off to fresh woods, in which he can
burn enough vegetation to manure his ])atch of rice or maize. Buddhism
is professed by the whole tribe, and their Lamas are all from Tibet; but
against the more actively malevolent spirits the aid of the Bijua or Ojha
is invoked. Their religion is very much that of the Limbu, behind a
veil of Buddhism of the Himalayan type. The Tibetan strain is much more
marked in the Murmi than in most of the tribes hitherto mentioned;
indeed, the usual name for the tribe is Tamang Bhotia, and the sub-
divisions are almost all Tibetan in their titles. The Miirmi have been long
in their present locality, and have half-assimilated a good deal of Brah-
manism which is obscuring the Buddhism they brought with them. But
though the Brahman officiates for them at the festivals of his creed, and
the Lama is called in for marriages, stones, trees and village gods are
not neglected, and if a Lama be not at hand, their worship is carried on
by any layman who has mastered the procedure. They rank as a pure
caste in Nepal, but will eat with the Kirata and Lepca. The majority of

138 5- Ethnography.

those enumerated in British territory are probably labourers in the tea gar-
dens of Darjiling. In their native place the Murmi are an agricultural class.

The Ncvar, of whom a few thousands arc found in the same locality
as the Murmi and Kirata, are not a caste, but the aggregate of the early
inhabitants of Nepal, differentiated into functional divisions which gradually
grew into castes. The Nevar are both Brahmanists and Buddhists, the
latter are attracted to the Tibetan frontier, whilst the others are gaining
ground on the south ranges and valleys. The two stand absolutely aloof
from each other in all social matters. The Nevar in British territory, being
away from the strict organisation imposed upon the community by the
Chief of the race ruling before the Gorkha, grow very lax in the matter
of intermarriage, and thus lose position if they venture back into their
native land.

§ 102. The five principal tribes of Nepal, known as the Mukhya, are
the Khas, the Gurung, the Mangar and the Sunuvar. It was the
combination of these which overthrew the Nevar rule in the middle of
the 1 8th century, and established that of the Gorkha. The Khas is a
thoroughly Brahmanised community, with a strong admixture of Brahman
blood. On the advent of the Muslim, many Brahmans had to fly for refuge
to the hills, where they settled amongst the local tribes and proceeded
to bring them into conformity with their own scheme of life. To help on
this task the families of highest rank were dubbed Ksatriya, and the same
rank was stipulated for by them for the offspring of their own order by
the hill women. These two stocks furnished the now dominant class in
the State, with the peculiarity that with K.satriya rank the patronymic
titles are all Brahmanic, from the caste of the father. It is also on record,
however, that in the 14th century, a-Rajput Chief of north Bihar dispossessed
an ancient Hill Rajput dynasty, and that the Gorkha Chief who in turn
dispossessed the intruder from the plains, was himself a direct descendant
of one of the Udepur line, who fled to Gorakhpur after defeat by the
Muslim, and set up a princiiiality of his own on the upper Gandak. Thus,
whilst the Aryan strain is undoubtedly existent in the Khas, the Mongoloidic
origin is no less apparent. The Gurung rank next to the Khas among
the fighting, or Gorkha, tribes. In their case there is no question of mixed
origin. Since, however, the Gurung has abandoned Buddhism for the creed
of his rulers, there has been, as between this tribe and the four others
of the Mukhya, not exactly an interchange of brides, but the condonation
of the abduction of them from each other. In the tribal worship and
ceremonial there remains a good deal of the Himalayan animism, imported,
probably, from the interior, and a member of the Lama sub-caste, though
not a professional ministrant, is often substituted for the Brahman, when
there is a suspicion of sorcery or witchcraft. The Mangar and Sunuvar
both hail from western Nepal, and both made their way east-ward by the
same route. Their appearance and the nomenclature of their subdivisions
stamp them as Mongoloid of the Tibetan type, though both are now what
are called "undeveloped" Brahmanists, like the rest, and are served by
Upadhya Brahmans, who suffer no degradation thereby. Both are agri-
culturists and soldiers, the Mangar also doing something in the \vay of
petty trade. In connection with the recruiting of so-called Gorkha soldiers,
mentioned at the beginning of this paragraph, it should be noted that the
term Gorkha is used outside the State of any recruit of a Nepal tribe,
but it correctly appertains to the Mukhya tribes only. At the same time.

i

Castes and Caste-Groups. G. Muslim Race Titles. 139

the Nepal rulers have for a generation or more taken into their service
recruits from the Kirata tribes, but they are brigaded into regiments by
themselves. In the British army, some of the Gorkha battalions contain a
good many of this class, especially in Assam.

G. Muslim Race Titles.

§ 103. Of the total Muslim population of India nearly 58 per cent bear
the titles of races foreign to the country. Those whose names imply Arab
descent amount to 30,442,000. About 4,239,000 nominally belong to the
tribes on the north-west frontier, and the remaining 434,000 affiliate them-
selves to races introduced by the Central Asian dynasties which successively
ruled from Delhi. It must not be supposed, however, that the proportion
of foreign blood is that indicated 1)y the prevalence of the above titles,
except in the case of the frontier races, who have naturally overflowed
into Sindh and upper India. On the contrary, in some parts of the country,
it is said that converts from Brahmanism are so deeply imbued with the
notion of a fourfold division of society, fostered by the traditional sacer-
dotal partition of the Indian world into Brahman, K.satriya, Vaisya and
S'udra, that they consider themselves bound, when acce|)ting Islam, to
enrol themselves as either Sekh, Saiyad, Mughal or Pathan. In eastern
Bengal, accordingly, the name of Sekh is practically assumed to connote
native, instead of foreign, origin. In the Panjab, again, and the region
round Delhi, the long supremacy of the ^Mughal has endowed that race
with a halo which is still attractive to the local convert. Nearly nine tenths
of the Turk, too, belong to a subdivision of Banjara, which, as already
stated above, adopted Islam en masse; and, finally, nearly all the Arabs
of Sindh bear the title of Kalhora, the ruling race before the Talpur. The
figures now to be reviewed, then, must be taken with the above qualifi-
cations. The latter, it will have been seen, apply most extensively to the
communities purporting to belong to the native land of the Prophet, which
ought to be the most honourable, as they are the most numerous on
the record.

§104. (a) Arabian (25,441,900). The small number returning themselves
as Arab, without detail, might be still further reduced were the Kalhora to
be treated as an indigenous body, bringing the total down to about 75,000.
The returns of Mother-tongue would justify still further diminution, but in
the west of India, where the Arab is chiefly found, the community is divided
into the Vilayati, or foreigners, principally from Hadramat, and the Muvallad,
or native-born, the latter being the progeny of Arab or sometimes Makrani
fathers by wives taken from some local Sunni caste, in whose household
the current vernacular is Hindustani. The Arabs settled in India perma-
nently are generally guards in the service of native Chief or kept by the
principal bankers in the same capacity. The others, true to their secular
connection with India, are merchants and traders, with the modern addition
of horse-dealing, in connection with the ports on the Persian Gulf The
two small tribes of Hans and Khagga, in the Panjab, are also said to be
Arabs who came by land and settled north of Multan. They are now
apparently merged in the Pathan or Jat tribes. The title of Sekh is
widely spread over the country, and, except in the Panjab and Kashmir,
predominates more or less over all Muslim designations. The common
practice just referred to, of taking this name on conversion is justified by

MO 5. Ethnography.

the Hadith, or saying of the Prophet "All converts to my faith are of me
and my tribe". In Lower Bengal, from which 80 per cent of the .Sckh
are returned, this title covers 85 per cent of the total Muslim population^.
In the Muslim -State of Haidarabad, the corresponding proportion is 70 per
cent, and in Mysore, also under rulers of this creed once, it is over
60 per cent. Elsewhere it ranges from 25 to 40. It is smaller, as is to
be expected, in upper India where Islam was the State religion, and in
the Panjab where conversion does not affect caste or social position, and
where, as in the upper Gangetic region, the larger communities often
contain a Brahmanic and a Muslim branch, giving the convert the oppor-
tunity of retaining his former status, with a change in his worship only,
and often a very slight one even in that. In Bihar, a jirovince which stands
between the ignorance of eastern Bengal and the cxclusiveness of the
upper valley, it is only the converts of the higher castes, such as Rajput,
Babhan or Kayasth, who are allowed to pass directly into a race-title.
Those of humbler origin have to spend a time in the probationary grade,
as it were, of Nau-Muslim, or raw-recruits, and their further advancement
depends upon their conduct or worldly prosperity.

The Sekh are much subdivided, though throughout the greater part
of India the sections have little more significance than the main title.
Originally, amongst the Arabs, the term denoted eldership or a position
of authority only. It subsequently became the special designation of the
Qurc.s, the tribe to which the Prophet belonged, and of the descendants of
his own family and of his relations. Thus, the Banu 'Abbas .Sekh are derived
from his uncle, 'Abbas; the Ha.simi, from his great grandfather; J'afari,
from his cousin. The .Siddiqi are so called from the first Khalif. Abu Bakr,
named As Siddiq, or the Truthful One. The second Khalif, Omar, was
called Faruq, the Distinguisher of Right from Wrong, and from him come
the Faruqi. The Ansari, or Helpers, were the inhabitants of Al Medinah,
who sheltered the Prophet, and so on with several more of these sub-
divisions. In some Provinces the details of Sekh have been tabulated,
but for the most part the value of the return is vitiated by the prepon-
derance of those who failed to have this information entered against their
names. At best, except in the north, the return indicates in most cases no
more than the personal preference of the householder. In the Gangetic
region, so far as the information goes, the favourite section is decidedly
the SiddlqT, and after it, the Qure.si. In the Panjab, too, and in Sindh the
Qure.si have been separately given, but the return is only partial.

The Saiyad, a title said to be derived fromSud, gain, are, strictly
speaking, the descendants of 'Ali, cousin of the Prophet, who became his
son-in-law, and the line is generally limited to his offspring by Fatma, not
by his other wives. Thus the primary division of the Saiyad is into the
claimants through Hasan and those through Hussain, the proto-martyrs of
the faith, but many call themselves after other relatives of the Prophet,
using the same titles as the Sekh. Others have adopted geographical
names, such as Bukhari, Sabzawan, BilgramT, Barha, the two last being
descended from a celebrated Saiyad of 'Iraq, whose family settled in upper
India, like many others, in the train of one of the Muslim conquerors.
Probably in all the tracts surrounding Delhi and the principal seats of
Muslim authority there are families of Saiyad who hold their estates by
inheritance from ancestors who rendered distinguished service to the Mughal
power either in the field or in administration. Indeed, one family is said

i

p

Castes and Caste-Groups. G. Muslim Race Titles. 141

to have "made four Timurides emperors, dethroned and killed two, and
blinded and imprisoned three". The genealogy of most of the Saiyad of
India, however, is not so well attested, and, apart from the selection of
this rank by converts of high Brahmanic caste, which is a practice said
to have received the approval of the great Emperor Akbar, it is reported
to be not uncommon for a Muslim changing his sect from Sunni to Si'ah,
to signify his belief in the rank of 'All as jjremicr Khalif, by adopting
himself into the company of the Apostles. Nevertheless, far down to the
south, there are Saiyad settled whose forefathers followed the fortunes
of some one or other of the Muslim invaders, and who now, though in
some cases reduced to take to lowly occupations for a living, generally
hold to their rank and intermarry only with other Saiyad, or members of
the Mughal or Pathan races, and occasionally, but as seldom as possible,
with some respectable local family of the same sect. For there are, it
should be noted, Sunni Saiyad as well as those of the Si'ah sect, to which,
in theory, all ought to belong. In the western Panjab the Saiyad is usually
a religious teacher, irrespective of race or descent, and too often is a
member of "that pestilential horde of holy men, who not only prey upon
the substance of the people but hold them in the most degrading bondage".
"The Pathan is a bigoted Sunni, yet he maintains more Saiyad than the
Baluc, once known as "the friends of 'AH".

§ 105. (b) Mongol (394,600). Of the two races which entered
India with the Ghaznavides and later, the Turk and the Mughal, it
is hard to say which is the more unduly magnified in the Census returns.
The inclusion among the former of the Turkiya sub-caste of Banjara
has been mentioned. Then, too, in Bihar and round Delhi, Turk is
the equivalent amongst the peasantry for any official, especially if he
be of the creed of Islam, and Mughal serves the same purpose in Orissa
and the east Dekkan. The real Turk in the north is the traveller or
merchant from Turkistan, who is a temporary sojourner in Kashmir and
Peshawar. The only permanent colony is that left by Timur in Hazara at
the end of the 14th century'. In the west coast, in Bombay and a few other
towns, and in Haidarabad, there are probably a few families of Osmanli.
The Mughal element, in the south and east is better defined, as the con-
vert of those parts does not affect the title, and those who bear it are
probably correctly described, being as they are, the representatives of
families brought into Bengal and the south Dekkan by the semi-independent
Viceroys of Delhi. In the north there is the tendency already mentioned
to assume the title of Mughal on conversion or on rising in the world,
which is found in the parts of the Panjab where Islam predominates but
the Pathan influence is not supreme. Along the Jamna, however, there are
considerable numbers of true immigrants, settled upon estates conferred
upon their family by the Turk Emperors, from Babar downwards. The dis-
tinction between Turk and Mughal, however, is not in such cases very
clearly drawn, and subdivisions are returned which are common to both,
as, for instance, Turkman, Qizilba.s, and even Caghatai, the tribe of Babar.
As a rule, the Mughal and Pathan, assuming them to be of really pure
descent, are not considered, away from the frontier, at all events, as equal
in rank to the Saiyad and Sekh, and their position, consequently, depends
a good deal upon that of the family in its neighbourhood. In the interior,
too, there is a tendency to introduce endogamous subdivisions, or more
correctly, perhaps, to make existing sections endogamous. There is also,

142 5- Ethnography.

at the lower edge of these communities, a fringe of dependents who are
either bastards of the upper classes, as among the Rajputs, or have taken
the title of their employers and patrons on conversion. These do not
intermarry with the Mughal or better families of the Sckh. In the west
of India, in addition to the Caghatai, there is a considerable sprinkling of
Persian settlers and refugees, who go by the name of Mughal. They are
strict Si'ah and do not intermarry with Indian Muslim. Most of them have
engaged in trade. The Caghatai, on the other hand, have become almost
an integral part of the Muslim masses, apd are Sunni, with the cu.stoms,
language, and religious observances of their neighbours.

i? io6. (c) The Pathan and Baluc (4,287,000). If the hypothesis of the
identity of the Pathan with the Paktyes of Herodotus be true, as is now
generally believed, these tribes must have been from time immemorial
neighbours of India, and even occupants of some part of the territory
which is now included in that country. Some of them, again, were people
amongst whom Brahmanism found a favourable reception, and then. Buddhism,
the latter especially lingering long in these secluded valleys and on the
high road to India which passes near them. The Pathan, however, accepted
with equal zeal and devotion the exceedingly narrow and superstitious
form of Islam now current amongst them, and anything less like the mild
and tolerant character of the Indian Buddhist than the present temperament
and habits of the frontier men of nowadays can hardly be imagined. At
the same time, the Pathan, like all highlanders in the tribal stage, has
his charm in his virile independence and his strict observance of the
national code of hospitality and asylum, even towards an enemy — the great
solace of his life. It cannot be denied, however, that the epithet of
"faithless", universally appended to his name by those who have to deal
with him, is, like most of the proverbial sayings of the country-side, very
well deserved, by at all events the hillmen. Those who have settled in
the plains of the Panjab, even though within easy reach of their fellow
tribesmen of the highlands, are soon softened by their circumstances, and
the more they prosper the less respect they show for the hard life they
have left behind. In the interior of India there is no Province or State
without its quota of this race, and, no doubt, looking at the extent to
which soldiers of fortune were settled by their victorious employers upon
the land overrun by them, there is a good deal of real Pathan blood
disseminated amongst them, but not to anything like the amount indicated
on the face of the returns for regions like Bengal or the peninsula. In
the former, indeed, the title of Pathan is regarded as the right of a con-
verted member of a Brahmanic military caste, and the further detail of
selecting a tribe or clan presents no more difficulty to him than that of
a Rajput clan does to an aspiring Kol.

The term Pathan is now used to denote any one speaking the Pakhtun
language, or Pastu, and thus includes the Afghan, a foreign race which,
however, has impressed its name upon the whole country. The Afghan,
whose Jewish origin is insisted on by several authorities, and regarded
as unproved by others, first settled in the hill tracts of Ghor and Hazara.
Thence they descended upon the Helmand valley, which was in the occu-
pation of the Gandhari, a Pathan tribe expelled from the Peshawar valley
by one of the Scythian invaders. These people were dominated and then
converted by the Afghan, who finally intermarried freely with them. The
Gandhari, however, took the first opportunity of reverting to their former

i

Castes and Caste-Groups. G. Muslim Race Titles. 143

seat, where, under the names of Yusufzai, Mohmand, etc., they now reside.
The Afghans, by this time known as Tarin, Sirani, and Abdall, or Durrani,
remained round Kandahar until the i8th century, when they transferred
their headquarters to Kabul. The Ghilzai, a Turk tribe which is Pathan
but not Afghan, arrived across the Bamian from Ghor, like its predecessors.
After rendering great assistance to Mahmud of Ghazni on his raids into
India, the Ghilzai took possession of the country between Jellalabad and
Qal'at-i-Ghilzai, and have since spread east and west from that nucleus.
In addition to the GandharT just mentioned, the Paktyes contained, according
to ancient writers, the Aparytai, or Afridi: the Sattagydai, or Khatak,
and the Dadikai, or Dadi, all of whom are ascribed to an Indian origin.
Along with the Afghan, Ghilzai, the Scythic Kakar, the Waziri (said to be
Parmar Rajputs\ and a few Turk accretions brought down by Sabaktagfn
and his successors, these tribes constitute the Pathan of to-day. The terri-
tories occupied by the ancient people of that name, however, have been
much altered^ The Kakar nearly obliterated the Dadi in Sewistan; the
Khatak and Afridi were dispossessed by the Turk to a great extent. But
through the operation of intermarriage and the adoption by all of the
Pa.stu language, the whole has been welded into one nation, with the
usual fictions as to common descent to explain the fusion.

The modern Pathan inhabitants of upper India were first introduced
by the Lodi and Sur dynasties, and consisted chiefly of Ghilzai, who were
not Afghan, nor, at that time, Pathan. They were soon followed, however,
by large bands of other tribes, who were generously endowed with estates
by the Ghazni Chiefs and also by Babar, whose original army grew like
a snowball as he moved it across the hills to the plains of promise. The
tribes most numerously represented in this distribution were the Yusufzai,
the Orakzai, Lodi, Kakar and Karlanri. The tribal organisation gets weaker,
as is only to be expected, as the distance from the frontier increases,
and is scarcely to be found in its original form east of the Jamna, where
the Rohilla community, well known in history, is probably the best-knit,
as it is the most prosperous, of the larger settlements of this race. In
addition to the Pathan colonies and the converts arrogating to themselves
that title, there is a floating population of from 100,000 to 150,000 Powindah,
or itinerant traders of Pathan nationalit\-. They belong chiefly to the Ghilzai
tribes, though, owing to their nomad life, their connection with their kins-
folk is of the loosest. Large caravans assemble in the autumn to the east
of Ghazni, and march in armed bodies through the dangerous country of
the Waziri and Kakar, to the Indus at Dera Ghazi Khan. Here they deposit
their arms, leave their families encamped on the grazing grounds along
the river, under the guardianship of a detachment of their fighting men,
and wander off across upper India, often as far as Bihar, selling the goods
and horses they have brought from Kandahar and Central Asia. When
these have been disposed of, the Powindah act as pedlars on behalf of
merchants in the larger towns. In the spring they re-assemble on the
Indus, and wend their way back to Kandahar, dispersing from that centre
by their various routes through Herat and Kabul to the north. Some few
of the band engage in contract labour for the season. There are gangs,
also, but not belonging to the regular Powindah, which remain longer in
India, taking up work as it suits them, and usually afi'ecting tracts well
known for their prosperity and the unwarlike character of their population.
In these lush pastures their superior size and strength, added to their loud

144 5- Ethnography.

and gruff voices provide them with a living until they are moved on by
the police towards a region where those qualifications are sufficiently

familiar to fail to extort respect or alimony.

i; 107. Baluc. A line drawn from Dcra Ghazi Khan through the
Sulaiman range due west to Quettah demarcates approximately the Pathan
on the north, from the Baluc on the south; but the latter have advanced
considerably to the north of this limit in the Indus valley, and have also
established large colonies in upper and middle Sindh. The Baluc state-
ment of their origin is to the effect that .they belong to Aleppo, and were
expelled from Syria on sectarian grounds. They found their way through
Baghdad and Kirman to Makran, where they lived for many generations
before they occupied Khalat and the south Sulaiman hills, which they took
from the Pathan, A large section of their community was expelled from
BalucTstan in a tribal dispute, and settled in Sindh. Members of these
exiled clans joined with their kinsfolk of the plains in rendering assistance
to the Emperor llumayun, when regaining India after his expulsion. They
were rewarded with grants of land along the Indus, and have now spread
well up the Cinab and Satlaj valleys. The result of this movement is that
there are now more Baluc in Sindh and the Panjab than were enumerated
in their native country, where they are outnumbered by the Brahui. There
are many Baluc tribes, but the predominant section is the Rind, from
which most of the rest claim to be descended. The La.sari stands next
in rank, but according to the tradition of the others, it was treated as
the Ksatriya were treated by Paras'urama, and swept off the face of the
country, thereafter being known only in middle Sindh, and there in a
disjointed condition which has never been repaired. The Rind, too, colonised
a part of upper Sindh, but are not found to any great extent elsewhere
in British territory, outside British Balucistan. The tribes best represented
on the frontier and along the rivers are the Marri, with their hereditary
foe the Bughtr, of the hills, and the Mazari, Gurchani, Leghari, Lund,
Bozdar, and of course, the Rind itself Except in upper Sindh and the
Dcra Ghazi Khan district, the Baluc of British domicile do not keep up
in parti bus the characteristic tribal organisation so strictly observed in
their own country. As they get higher up the rivers, they tend to amal-
gamate with the Jat and Pathan. In the south-west Panjab, indeed, every
camel driver is called Baluc, owing to the marked addiction of the race
to that occupation. In spite of this dilution of the original stock, the
independence of the artificial restrictions of caste and the strongly-marked
character of the Baluc and Pathan alike, different as these peoples are
in other respects, have had very considerable effect upon the customs and
general tone of the population in the midst of which these races have
settled. This influence, according to competent observers, has been greater
than that of the political supremacy of Islam in producing that laxity in
religious matters which is generally attributed to the latter cause alone.
It should not be forgotten, however, that the people of the west enjoyed,
many centuries before a single Muslim was in existence, a unique repu-
tation in the eyes of the Singers on the Sarasvati, for religious indifference
and "neglect of rites", which justified their inclusion amongst the MIcccha.

§ io8. Brahui. Last among the more definite communities acknow-
ledging Islam is that of the Brahui, inhabiting Balucistan and Upper
Sindh, of whom only 48,000 were enumerated within the scope of this
survey. For centuries the Brahui have been Muslim, and have inter-

i

Castes and Caste-Groups. G. Muslim Race Titles. 145

married with Jat and Baliic, and have even admitted adult recruits from
these races into their trii^es. Nevertheless, they have preserved their
distinct physical features, being shorter and more swarthy than their neigh-
bours; and, though, as remarked in the Introduction, their language has
been overlaid with SindhT and Baluci, they keep, for domestic use at
all events, a tongue undoubtedly Dravidian in its main characteristics. In
common with their neighbours, from whom they have perhaps borrowed
it, they hold the tradition of Arab descent, Aleppo being their chosen seat
of origin. On the other hand, they are cijually certain that they have never
lived in any other country but that which they now occupy. Setting on
one side the conjecture that the Brahui are of Scythian race, for which
there is little corroborative evidence, it is known that there was of yore
a considerable Indian population settled along the hill-country west of
Sindh, with its own customs and temples. It is possible, therefore, that
the Brahui may denote the high-water mark of the Dravidian extension
northwards, left derelict and isolated under the protection of the desert,
after the Indus had changed its course and the tide of Aryan occupation
had absorbed the bulk of the darker race. In the present day the Brahuf
are specially addicted to the rearing and tending of camels. They enjoy
a good social position in Balucistan, but are rarely found far from their
wide pastures, except for purposes connected with their occupation.

With these tribes ends the list of the communities which have been
selected as representative of the different elements of which the vast and
complicated society of India is compounded. That the review of their
leading characteristics is imperfect has been fully admitted throughout,
and the certainty of error will not be denied by any one who has attacked
even the outworks of a task of this nature. It needs but little experience
of Indian life to bring home to the student of ethnography the vanity of
thinking that the whole field can be adequately surveyed in the light of
such knowledge as can possibly be acquired by a single individual. Here,
indeed, if anywhere, a little knowledge is dangerous, because, as has been
abundantly shown in the course of this review, Indian society differs from
tract to tract to an extent which inevitably involves the lurking danger of
being led astray by analogy or similarities of nomenclature, rites or customs,
into the assumption that what is true of a community in one part is equally
applicable to a body of perhaps the same name elsewhere. Information
upon such distinctions must be obtained, as a rule, at second-hand, and
fortunately, the supply thereof has greatly increased of late years both in
amount and quality and has received valuable additions even since the
body of this review was written. It is on such material that reliance has
been mainly placed in the attempt here made, perhaps rashly, to give a
word-picture of society as it exists to-day in India, not merely geographi-
cally, but as a whole.

In.lo-Arjan Research. 11. 5.

146

5- Ethnography.

APPENDIX A

Summary of Caste-Groups.

A. («} 24 — 31 ) Special Groups.

(§ 24—26) Brahman

(§ 27) Rajput . .

(§ 28 — 29) Traders

Banya unspec*

Agarval

Agrahari

S'rimali

Porval

Osval

HumbacI

Khatrl

Arora

Bhatia .

Lohana

Subarnabanik

Balija

KOmati

Banjiga

Vatluga

Cetti .

Khojah

Meman

Bohra .

Labbai

Mappila

Jonakkan

(§ 30) Writers
Khatri. . . .
Kayasth . . .
Prabliu . . .
Brahmaksatriya
Karan-Mahant .
Kanakkan
Karnam . . .
Vidhur . . .
Vaidya . . .

14,893,300
10,040,800

3,163,300
557,600

92,000
227,400

75,000
382,700

60,700
585,000
732,100

60,600
572,800
154,800
534,700
656,300
173,400

95,900
320,000
155,300
112,100
177,300
426,300
925,200
100,300

138,000

2,149,300

28,800

4,200

195,000

63,000

42,800

39,200

90,000

Atit .
Sadliu

Jogi ■
Faqir
And! .
iJasari
Panisavan

151,800
67,800

212,500
1,212,600

101,400
48,300
13,700

5. (§ 31) Religious Devotees.

r Gosai 152,600

I Bairagi 765,200

B.

(§ 32—53) The Village Com-

munity.

6. (a) (§ 33 — 34) Landh

olders.

Military etc.

" Jat

7,086,100

Gujar ....

2,103,100

Avan ....

686,000

Khokhar

117,500

Gakkhar . . .

30,000

Kathi . . . • .

27,400

Sumro ....

124,100

Sammo ....

793,800

^ Taga ....

165,300

Babhan-Bhuinhar

1,353,300

r Rajbahsi-Koc . .

2,408,700

L Ahum ....

178.000

Khandait

720,300

Maratha . . .

5.029,300

r Razu ....

113,500

L Velama ....

519.900

- Kalian ....

494,600

Maravan

350,000

Agamudaiyan . .

318,600

Nayar ....

1,046,700

Kodagu. . . .

36,200

(b) (§ 35—36) Peasants.

Kambo .... 183.600

Me6 395,000

Thakar 102,200

Rathi 39,300

Raut 81,900

Ghirath 170,100

Kanait 389,900

Appendix A. Summary of Caste-Groups.

147

(c)

" Kurmi . .

3,873,600

Kucri . .

1,784,000

Lodha . .

1,663,400

Kisan . .

442,700

" Kavar . .

iS6,ioo

Kolta . .

127,400

_ Kirar . .

166,700

~ Kalita . .

203,400

Halvai-Das

29,200

Kaibartta .

2,665,100

Sadgop . .

579,400

Casa . .

870,500

Gangauta .

82,600

POd\ . .

464,900

_ Namasudra

2,031,700

Kunbr .

2,700,000

Kanbi . .

1,350,600

. K6I1 . . .

2,477,300

Vakkaliga .

1,392,400

Lingayat unsp

d.

2,612,300

Pancamasale

431,100

Caturtha

111,600

Banta . .

1 20,600

_ Cauda . .

162,500

" Kappu-Reddi

3,110,200

Kamma . .

974,400

Telaga . .

644,200

Kalingi . .

126,900

Tottiyan

151,000

' VeOalan. .

2,464,900

_ Nattaman .

151,300

(§ 37) Gardeners etc.

■ Baraf 545,900

Senaikkudaiyan

39,300

. Kodikkal . .

60,000

Arain

1,026,500

. Maliar . .

159,900

■ Mali . . .

1 ,948,600

Kachi . . .

1,260,200

Murao . .

662,900

Saini . .

200,600

Tigala . . .

64,800

(§ 38) Cattle-breeders.

Ahir 9,841,900

Goala-Golla . . 1,357,400

Gaura 431,600

Rabari 253,900

Gh(")Si . .

58,500

Kannacjiyan . . 22,500

8. (i; 39) Artisans.

(a) Combined castes (Panckaisi)

Kammalan .... 644,600

Katfisala

295,500

Pancala . .

323,800

(b) Sonar . .

1,271,800

Niyariya .

18,700

(c) r Tarkhan .

754,500

L Barhai .

1,133,100

Sutar . .

581,100

Khati . .

219,400

(d) LOhar . .

1,605,100

Kamar . .

757,200

(e)- Raj . . .

26,000

. Thavi . .

2,300

Gaunili .

8,700

. Kadio . .

14.400

(f) Kasera . .

138,600

Thathcra .

57,800

Tambat . .

10,400

9. (§ 40) Weaver

' Patnuli . . .

. . 90,500

Patve . .

72,000

. Khatri . .

56,200

i~ Tanti . .

772,300

L Tantva . .

197,900

r Perike . .

63,000

Janappan .

83,000

Kapali . .

144,700

_ DhOr, . .

24,400

r Panka . .

726,700

j Ganda . .

277,800

L Dombfi .

76,400

Kori . . .

1,204,700

Julaha . .

2,907,900

L Balahi . .

585,100

Kaikkohn .

354,700

Sale . . .

639,300

Togata . .

64,500

Devanga

288,900

Neyige unsp"*-

97,000

Jug. . . .

536,600

Ko.sti . .

277,400

148

5. Ethnography.

(§ 41) Oil-presscrs.
Tclf-Ghanci . . . 4,060,300
Kalu 154,900

Vaniyan

Ganiga .

(§ 42) Potters.

Kumhar

Kusavan

{§ 43) Barbers

Nai-Nhavi . .

Hajam . . .

Ambattan . .

Marayan . .

Mangala . .

Bhandari . .

187,500
114,909

3,376,300
145,500

2,458,400
534,300
219,700
S,8oo
277,600
120,300

13. (§ 44) Washermen.
DhobT-Parit . . . 2,016,900
Vannan .... 253,200
Veluttcclan . . . 24,500
Agasa 122,200

_ Cakala 470,800

14. {§ 45) Fishers, Boatmen

and Porters.

' Mallah unsp''- . . 721,600

Patni 63,700

Tiyar 270,900

Malo 246,600

. Kevat 1,110,800

- Kahar 1,970,800

Dhimar 291,200

Jhfnvar 477,700

Machi 288,600

. Mohano .... 113,100

Bhof 169,800

Boya 530,400

Palle (about) . . . 150,000

Besta 230,400

Kabbera-Ambiga 76,500

Moger 38,200

^lukkuvan .... 20,400

S'embadavan . . . 54700

(§ 46) Stone, Salt and
Lime-workers.

Bind 219,700

Cain 158,600

Gonrhl 165,200

17-

16.

Luniya-Nuniya

807,400

Kharol . . .

12,700

Rchgar .

14,400

Kharvi .

50,000

Agria .

270,400

Uppara .

260,000

Uppiliyan

43.700

Patharvat

23,400

Baiti-Cunari .

iS.ioo

(§ 47) Toddy-drawers.

Pasi 1,408,400

Bhandari

176,000

Paik . .

80.900

Billava .

145,600

Tivan

580,000

Tandan .

19,000

Ijavan

791,100

.S'anan .

759.300

I.liga

337,400

Gaundia

361,500

Segidi .

53,700

Yata . .

52,700

§ 48—49) Field-labourers.

Dhanuk .... 804,200

Arakh 76,400

Dhundia-Dhodia . . 1 10,200

Dubla-Tala

via

141,800

BagdT 1,042,500

Baurl 705,600

Rajvar 166,400

Musahar .... 664,700

. Bhar 458,500

Dhakar 125,700

Palli 2,572.300

PaUan 836,500

Pulayan-Ceruman 524,500

Paraiyan .... 2,258,600

Mala 1,863,900

Holeya 866,200

Mahar 2,561.600

DhecJ 378,800

(§ 50) Leather- workers.

Camar .... 11,176,700

Megh 140.500

Dagi J 54.700

Appendix A. Summary of Caste-Groups.

149

19.

Madiga . . .
Mang

S'akkiliyan.
Jloci ....
Bambhi (about)

(§ 51) Watchmen

Barvala .
Gh5tval .
Kandra .
Ambalakkaran
Mutraca
Khangar
Mind . .
Dosadh .
i\I51 . .

Berad-Bedar
Ramos'i . .

i§ 52—53^ Scaven

Bhangi-Mihtar

Cuhra . . .

Mazbr (about)

Bhuinmalf . .

Hari and Kaora
L Haddi . . .

Dom ....
_ Ghasiya . . .

1,281,200
579,900
487,500

1,007,800
200,000

101,700
88,800
151,500
162,500
329,100
113,700
581,900
,258,200
145,700
646,000
60,800

839,200
1,329,400

38,000
131,600
306,500

28,100
855,600
1 19,300

54—58) Professions
Subsidiary.

(^ 541 Bards and Gc
nealogists.

Bhat . .
Bhatrazu

Raj-Bh5t
Caran .
Mirasi .

577,700
28,000
11,200
74,000

291,600

(§ 55) Astrologers etc.

Jo.si . .
Dakaut .
Ganak
Kanis'an
Panan .

Velan .
Garpagarl

83.700
15,600
20,500
15,700
33.300
27,700
8,800

23- (§ 56— 57)Temple-services.

(a) Priests.

r Pujari 880

L Bhojki 1,070

r Bhojak 1,200

L Scvak 6,800

r Pandaram .... 68,600

L Valjuvan .... 85,300

r Tambala .... 3,800

L Jangam 405,000

r Garuda .... 20,600

_ Bharai 66,000

Ulama 36,200

(b) Servants.
Phulari-Hugar 1 5,700

Gurao 94,000

Bari 89,600

r Satani 77,400

L Devadiga .... 23,800

24. (§ 58)Dancers andSingers.
Besiya, Kancan etc. 5 7. 700
Kalavant .... 20,000

~ DasF-Dcvali . . 25,300

L Bogam 32,900

D- (§ 59—68) Urban Castes.

25. (§ 60) Grocers etc.
Attari . .
Gandhabanik

r Kasarvani .
L Kasaundhan

Gandhi . .

Kunjra . .

Tamboli

26.

27-

(§ 61) Grain-parch

Confectioners.
Bharbhunja
Bhathiara ....

Kandu

Halvai . . . . :

Mayara

Godiya-Guria .

(§ 62t Butchers.

Qasab

Khatik

5,900

141,100

79,700

99,700

3,700

285,400

209,500

ers and

359,500
58,200
667,900
260,000
149,200
150,400

369,500
332,300

ISO

5- Ethnography.

28. (§ 63) Pedlars and Glass-

workers.

Bisati 3,600

Ramaiya .... 5.300

Manihar .... 102,300

Curihar 55, 500

Kancar 19,100

. Lakhcra . . . . 60,100

Gazula 102,000

Patra 61,400

S'ankhari .... 14,800

29. (§ 64—67) Artisans.

(a) Tailors.

Darji 831,100

S'impi 36,800

(b) Dyers etc.

Chipi 269,400

Bhausar 38,200

Rangrej 137,000

NilarT 48,300

Galiara 1,100

(c) Cotton-scutchers.

Pinjari 50,800

Behna 362,500

Dhuniya .... 272,800

Dudekula .... 74,500

(d) Distillers and Liquor-

seile rs.

Suriri-S'aha . . . 724,800
Kalal-Kalvar . . . 1,000,200

30.

(g 68) Domestic

Servants.

Bihisti 107,500

Gola . . .

39,700

Kuta . .

6,400

Cakar . .

163,600

Khavas . .

30,600

S'udra . .

285,000

Sagirdpesa

47,100

Parivaram .

18,900

31-

E. (§ 69—79) Nomads.

(§ 69) Carriers.

Banjara 496,400

I.abana 349, 500

Thori .
Pcndhari

41,800
10,100

32.

33-

34-

35-

36.

37-

38.

39.

(i; 70) Shepherds and

Wool-workers.
Gaddi 103,800

Garlariya . .
Dhangar-Hatkar
Kurubar
Idaiyan . . .
Bharvad . .

1,272,400
1,015,800
1,068,000
702,700
102,900
{§ 71) Earthworkers.
Od-Vaddar . . . 903,100

Bcldar 214,700

Kura-Khaira . . . 166,500
(§ 72) Knife-grinders etc.
S'ikligar .... 21,000

Ghisadi 8,400

Khumra . . . . 1,100

Takari 6,500

(§ 73) Bamboo-workers.

Turi 68,000

Basor-Barisphora . 96,000
Burud-^Iedar . . . 87,600
Dharkar .... 43,500
(§ 74) Mat and Basket-
makers.

Kanjar 34,000

Kuravan-Koraca 234,800

Yerukala .... 65,500
Kaikadi .... 14,200
(§ 75) Mimes etc.
Bahurupiya . . . 3,900

Bhand 10,600

Bhavaio .... 6,000
GundhaH .... 27,500
(§ 76) Drummers etc.

Dafali .
Nagarci
DhOli
Bajania ,
Turaha

50,200
20,600
43.700
14,400
77,300

(§ 77^ Juggles, Acrobats,
Snake-charmers etc.

Nat 162,300

Bazigar 27,000

Appendix A. Summary of Caste-Groups. 151

Dombar-Kolhati . . 39,400

" Kand 612,500

Gopal 7.100

Kondu-Dora . . . 88,700

40. (§ 78) Thieves etc.

Poroja 91,900

Gadaba 41,300

Bagariya .... 30,900

Jatapu 75,700

Bediya 57,500

_ Savara 367,400

Habura 4.300

(b) (§ 87) Western Belt.

Bhamtiya-UcH . . 6,100

Kurku-Korva . . . 181,800

41- (i;79)Hunters and Fowlers.

Bhil . . .

1,198,800

' Bavariya-Mogl
Aheriya . .
Baheliya .

liy:

1

30,300
35,400
53,600

BhilSla .
Dhanka .
Tadvi .

144,400
66,ioo

10,500

Mahtam

82,900

Nihal

6,900

Sahariya

1 36,400

Gamta .
- Patelia .

49,300
91,000

1 Vaghri . .
L PardhI . .

114,000

32,000

Naikada

90,200

Nayak

25,100

Veilan . .

25,500

Chodra .

!;8.200

Valaiyan .

383,000

Vettuvan .

74,900

(c) (§ 88) Sahyadri.

!_ Kuriccan .

9,600

Katkari 93,000

Vfirli . .

. . 152,300

F. (§ 80—102) Hill Tribes.

Ghat-Thakur

. . 122,300

42. (a) (§81—86) Central Belt.

id) (§ 89-90) N

Igiri etc.

K61 299,000

~ Kuruman .

. . 10,600

Ho . . .

385,100

Irula . .

. . 86,100

Munda .

466,700

Toda

. . 800

Bhumij . .

370,200

_ KOta

■ 1,300

Bhuinya

789,100

r Kanikkan

. • 4,100

Kharvar

1 39,600

L Malaiyan

. . 11,200

Baiga

33,900

r Yanadi .

. . 103,900

Ceru

30,200

L Cencu .

. . 8,300

Kharia

120,700

Santal

1,907,900

43- t§ 91 — 100) Assam Tribes.

. Mahili

66,800

(a) Bodo-Kacari . . . 242,900
Garo 162,200

r Birjia

5,700

L Juang

11,200

Lalung 35,500

Uraon

614,500

Rabha 67,300

Male

48,300

Mec 99,500

. Mal-Paharia

35,000

Hajong 8,800

" GOnd . .

2,286,900

: Tipparah-Mrung 111,300

Majhvar

52,400

L Cutiya 85,800

Bottada-Bhatr

a

50,100

Halaba . .

90,100

(bl Miri 46.700

Pathari . .

2,900

Abor 320

Pradhan

22,900

Daphla 950

- KOyi .

115,200

Aka . .

28

152 5- Ethndc.raphy.

(c) Khasi ui,6oo

44. (§ loi — 102) Himalayan Ne-

Sainteng

47,900

pali) Tribes.

(d) Mikir . .

87,300

" Khambu .... 46,500

Yakha .

2,400

(e) Naga unsp"'-

78,900

_ I.imbu

24,600

Angami-Teng

ma

27,500

I-epca

18,000

Ao . . .

26,800

Murmi

33,900

Sema-Sima

4,700

Nevar

11,500

Lhota . .

19,300

, "" Khas

15,900

Rengma

5,600

Guriing

16.600

(f) Kuki unsp""-

67,200

Mangar

23,900

Meithci . .

69,400

Sunuvar

6,900

Lu.sci . .

63,600

,. Gorkha unsp^"-

18,400

(g) San unsp''-

1,850

G. (§ 103—108) Muslim Race Titles.

Khamti . .

2,000

45. (a) r Arab unsp''- . . 96,700

Phakial . .

220

; Sekh . .

23,836,800

Nora . .

140

[ Saiyad

. 1,508,400

Turung . .

400

(c) r Turk

5.700

Aiton

80

L Mughal

388,900

[Ahom*

178,000]

(d) r Pathan

3,204,500

(h) Singpho

800

Baluc

1,034,300

Doania . .

1,000

\_ Brahuf

. . 48,200

Included amongsl Landed-Military in 6 (a).

Total of selected Castes and Tribes 265,701,200.

Appendix B. Caste Index.

153

APPENDIX B.

Caste Index.

Caste

Group

1 . 0 c a 11 1 y

Abor

43(b). Hill tribe

Assam Himalaya

Agamudaiyan

6(a). Landed-dominant

Tamil

Agarval

3. Traders

North and West

Agasa

13. Washermen

Karnatic

Agrahari

3. Traders

Agra

Agria

15. Saltworkers

Agra and West Coast

Aheriya

41. Hunters and towler.s

Panjab and Agra

Ahfr

7. Cattle-breeders

Upper and Central India

Ahom

6(a). Landed-dominant

Assam

Alton

43(g). Hill tribe

E. Assam

Aka

43(b). Hill tribe

Assam Himalaya

Ambalakkaran

19. Watchmen

Tamil

Ambattan

12. Barbers

Tamil

Ambiga = Kabbera

Andi

5. Religious mendicants

Tamil

AngamT-Tengima

43(c). Hill tribe

E. Assam

Ao

43(e). Hill tribe

E. Assam

Arab

45(a). Muslim race

Panjab and West

Arain

6(c). Market-gardeners

Panjab

Arakh

17. Field-labourers

Agra etc.

Arura

3. Traders

W. Panjab

Atit

5. Devotees

Bengal and North

Attari

25. Perfume-makers

North and Centre

Avan

6(a). Landed-dominant

Panjab

Babhan-Bhuinhar

6(a). Landed-dominant

Ganges Valley, Bihar

Bagariya

40. Thieves

Cent. India

Bagdi

17(a). Field-labourers

Bengal

Baheliya

41. Fowlers

Panjab

Bahuriipiya

37. Mimes

Panjab and Ui)per India

Baiga

42(a). Hill tribe

Cent. Prov.

Bairagi

5. Devotees

Univers. N. and Centre

Baiti

15. Lime-burners

Bengal

Bajania

38. Drummers etc.

West

Balahi

9. Weavers

Rajputdna etc.

Balija

3. Traders

Telingana

Baluc

45(c). Muslim race

Panjab and Sindh

Bambhi

i8. Shoemakers

Rajputana

Banjara

31. Carriers

North and Centre

«54

5. Ethnography.

Taste

(>r'iu\i

I . (J c a 1 i t y

15anjiya

3. Traders

Karnatic

Bahsphura-Basor

35. Bamboo-workers

1 Upper and West. India

Banta

6(b). Peasants

j Kanara

Banya unsp<*-

3. Traders

Univ. except in South

Barai

6(c). Bctel-vine-growcrs

Univ. except in South

Barhai

8(c). Carpenters

Upper India

Bari

23(b). Leaf-plate-makers

Upper India

Barvala

19. Watchmen

Panjab

Bas6r = Baiisphora

Bauri

6(c). Field-labourers

Bengal

Bavariya

41. Fowlers etc.

Panjab and Agra

Bazigar

39. Acrobats etc.

Panjab

Bcdar = Bcrad

Bediya

40. Disreputable nomads

Upper India

Bchna

29(a). Cotton-scutchers

Upper India

Beldar

33. Earth- workers

North and Centre

Berad-Bedar

ig. Watchmen

Karnatic

Bi'siya-Kancan

24. Dancers and singers

Upper India

Besta

14. Fishermen

Telingana

Bhand

37. Mimes

Panjab etc.

Bhandari

12. Barbers

Orissa

Bhandari

16. Toddy-drawers

West Coast

Bhangi-Mihtar

20. Scavengers

All but in South

Bhar

i7(ai. Field-labourers

Bchar etc.

Bharai

23(a). Shrine priests

Panjab

Bharbhunja

26. Grain-parchers

Upper India

Bharvad

32. Shepherds

West

Bhat

21. Bards and genealogists

Upper and West. India

Bhathiara

26. Public cooks

W. Panjab

Bhatia

3. Traders

West

Bhatra = Bottada

Bhatrazu

21. Bards and genealogists

Telingana

Bhausar

29(b). Calenderers

West

Bhavaio

37. Actors

West

Bhil

42(b). Hill tribe

West Belt

Bhilala

42(b). Hill tribe

West Belt

BhOi

14. Fishers and porters

Dekkan and West

Bh6jak

23. Priests to Jains

Rajputana

BhOjki

23. Priests of hillmen

Panjab

Bhiiinhar = Babhan

Bhiiinmali

20. Scavengers

Bengal and Assam

Bhuinya

42(a). Hill tribe

Bengal and Cent. Belt

Bhumij

42(a). Hill tribe

Bengal

Bihisti

30. Water bearers

North and Centre

Billava

16. Toddy-drawers

Kanara

i

Appendlx B. Caste Index.

'55

Caste

Group

Locality

Bind

15. Stone and lime-workers

Bihar and Oudh

Birjia

42(a). Hill tribe

Bengal

Bisati

28. Pedlars

Panjab etc.

Bodo = Kacari

Bogam

24. Dancers

Telingana

Buhra-Vohora

3. Traders and cultivators

West

Bottada-Bhatra

42(a). Hill tribe

South Cent. Belt

Buya

14. Fishers etc.

Telingana

Brahmaksatriya

4. Writers

Gujarat

Brahiij

45. Muslim race

Sindh Frontier etc.

Ruriitl-Mcdar

35. Bamboo-workers

Dekkan and Karnatic

("ain

15. Stone-workers

Oudh and Bihar

Cakala

13. Washermen

Telingana

Cakar

30. Domestic servants

Rajputana

Camar-Khalpo

18. Leather-workers

Univ. except in South

Caran

21. Genealogists

West

Casa

6(b). Peasants

Orissa

Caturtha

6 (b). Cultivators and traders

Karnatic

Cencu

42(d). Hill tribe

Eastern Ghats

Ceru

42(a). Hill tribe

Bengal

Ceruman = Pulayan

Cetti

3. Traders

Tamil

Chipi

29(b). Calenderers and dyers

Upper India

Chodra

42(b). Hill tribe

West

Cuhra

20. Scavengers

Panjab

("unari-Baiti

15. Lime-burners

Upper India and Bengal

Ciirihar

28. Pedlars and glass-workers

North and Centre

Ciitiya

43(a). Hill tribe

Assam

Dafali

38. Mendicant drummers

Agra and Bihar

Dagi

18. Leather-workers

Panjab Hills

Ddkaut

22. Astrologers

Agra etc.

Uaphla

43(b). Hill tribe

Assam Himalaya

Darji

29(a). Tailors

Universal

Dasari

5. Devotees

Telingana

Dasi-Devali

24. Dancers

Telingana and Karnatic

Dcvadiga

23 (b). Temple servants

Telingana and Karnatic

Devali ^ DasI

Devahga

9. Weavers

Karnatic

Dhakar

17. Field-labourers

Rajputana etc.

Dhangar-Hatkar

32. Shepherds

Dekkan

Dhanka

42(b). Hill tribe

West Belt

Dhanuk

17. Field-labourers

Agra and Rajputana

Dharkar

35. Bamboo-workers

Agra and Rajputana

1 56

5. Ethnography.

Caste

I,(.kalit\

bhcM.I

Dhimar

Dhobf-Parit

Dhoriia = Dhunc.lia

Dholi

Dhor

Dhuldhoya = Niya-

riya
Dhundia-Dhodia
Dhuniya
Doania
Pom-Dumna
Dombar-Kolhati
Domba
Dosadh
Dubla-Talavia
Dudekula
pum = Mirasi
Dumna = Dom

Faqir

Gadaba

Gadariya

Gaddf

Gakkhar

Galiara

GamaHa = Gaundla

Gamta

Ganak

Ganda

Gandhabanik

Gandhi

Gangauta

Ganiga

Garo

Garpagari

Garuda

Cauda

Gaundi

Gaundla-Gamajla

Gaura

Gazula

Ghanci = Teli

Ghasiya

17. Village menials
14. Fishers etc.
13. Washermen

38. Drummers
9. Hemp-weavers etc.

17. Field-labourers
29(c). Cotton-scutchers
43(h). Bastard Singpho
20. Scavengers
39. Acrobats etc.
9. Weavers
19. Watchmen
17. Field-labourers
29(c). Cotton-scutchers

5. Religious mendicants

42(a). Hill tribe
32. Shepherds
32. Shepherds
6(a). Landed-dominant
29(b). Indigo-dyers

42(b). Hill tribe
22. Astrologers

9. Weavers
25. Grocers
25. Grocers

6(bi. Peasants
10. Oil-pressers
43(a). Hill tribe
22. Hail-averters
23 (i). Low priests

6(b). Peasants

8(e). Masons
16. Toddy-drawers

7. Cattle-breeders
28. Pedlars

20. Scavengers

West

Ujjper and Cent. India

Univ. cxce|)t in South

West
Dekkan

West

North

Assam

Upper India

E)ekkan

N. E. Madras

Bihar

West

Telingana

Universal

N. E. Madras
Upper India
Panjab Hills
Panjab
West

West

Assam

East Cent. Prov.

Bengal

Dekkan etc.

Bihar

Ivarnatic

Assam

Cent. Prov.

West

Karnatic

Dekkan

Telingana

Bengal

Telingana

Ganges Valley

i

I

Appendix B. Caste Index.

157

Caste

Group

Locality

(Ihat-Thakiir

42(c). Hill tribe

Sahyadri

(Ihatval

19. Watchmen

Bengal

Ghirath

6(b). Peasants

Panjab Hills

Ghisadi

34. Knife-grinders

Dekkan

(.hosi

7. Cowherds

Upper India

Goala-Golla

7. Cattle-breeders

Upper India

Gudiya-Guria

26. Confectioners

Bengal-Orissa

G61a

30. Rice-pounders

West and North

Golla = Goala

Gonc.l

42(a). Hill tribe

Cent. Prov.

Gondhali

37. Ballad-singers

Dekkan

Gurirhi

15. Stone-cutters

Bihar and Oudh

Gopal

39. Jugglers

Dekkan

Gorkha unsp''-

44. Himalayan tribe

Nepal

Gosai

5. Devotees

Univ. except in South

Giijar

6(a). Landed-dominant

Panjab and Agra

Gurao

23(b). Temple-servants

Dekkan

Giiria = Gocjiya

Giiriing

44. Himalayan tribe

Nepal

Habura

40. Thieves

Upper India

Haddi

20. Scavengers

Orissa

Hajam

12. Muslim barbers

Universal

Hajong

43(a). Hill tribe

Assam

Halaba

42(a). Hill tribe

S. E. Cent. Prov.

Halvai

26. Confectioners

Upper and East. India

Halvai-Das

6(b). Peasants

Assam

Hari-Kaora

20. Scavengers

Bengal

lliukar = Dhangar

HO

42(a). Hill tribe

Bengal

Holeya

17. Village menials

Karnatic

Hugar = Phulari

Hiimbai.l

3. Traders

West

Idaiyan

32. Shepherds

Tamil

Tdiga

16. Toddy-drawers

Telingana

llavan

16. Toddy-drawers

Malabar

Iriila

42(a). Hill tribe

Nilgiri etc.

Janappan

9. Hemp-weavers

Tamil

Jahgam

23(a). Liiigayat priests

Karnatic [putana

Jat

6(a). Landed-dominant

Panjab, Agra and Raj-

Jatapu

42(a). Hill tribe

N. E. Madras

Jhinvar

14. Fishers and water-bearers

Panjab

J'"'gi

5. Devotees

Upper India

1 58

J. Ethnography.

( asti-

Jonakkan

( I r o u 1 >
3. Traders

I.iM alit\
Malabar

Josf

22. Astrologers

Univ. except in South

Juang

42(a). Hill tribe

Orissa Hills

J"g'

9. Weavers

Bengal

Julaha

9. Weavers

U|)i)cr India

Kabbcra-Ambiga

14. Fishers

Telingana and Kanara

Kacari-Bodo

43(a). Hill tribe

Assam

Kacf

6(c). Market-gardeners

I'jiper and Central

Karliu

8(e). Masons

West

Kahar

14. Fishers and porters

Upper India

Kaibartta

6(b). Peasants

Bengal

Kaikacli

36. Mat-makers

Dekkan

Kaikkujan

9. Weavers

Tamil

Kalal-Kalvar

29. (d) Distillers

Upper and Cent. India

Kalavant

24. Dancers

West

Kalirigi

6(b). Peasants

Telingana

Kalita

6(b). Peasants

Assam

Kalian

6(a). Landed-dominant

Tamil

Kalu

10. Oil-pressers

Bengal

Kalvar = Kalal

Kamar

8(a). Blacksmiths

Bengal

Kambo

6(b). Peasants

Panjab

Kamma

6(b). Peasants

Telingana

Kammalan

8(a). Artisans

Tamil

Karhsala

8(a). Artisans

Telingana

Kanait

6(b). Peasants

Panjab Hills

Kanakkan

4. Writers

Tamil

Kanbi

6(b). Peasants

West

Kancan = Bcsiya

Kancar

28. Glass-workers

Upper and Cent. India

Kand

42(a). Hill tribe

N. E. Madras

Kandra

19. Watchmen

Orissa

Kandu

26. Confectioners

Univ. except in South

Kanikkar

42(a). Hill tribe

Malabar

Kanisan

22. Astrologers

Malabar

Kanjar

36. Mat-makers

Upper India

Kannadiyan

7. Cattle-breeders

Tamil

Kaora = Hari

Kapali

9- Jute-weavers

Bengal

Kapu-Reddi

6(b). Peasants

Telingana

Karan-Mahant

4. Writers

Orissa

Karnam

4. Writers

Telingana

Kasar-Kascra

8(f). Brassmiths

Univ. except in South

KasarvanI

25. Grocers

Agra and Oudh

Appendix B. Caste Index.

159

Caste

Ci r 0 u 1 1

Local it \

Kasaundhan

25. Grocers

Agra and Oudh

Kasera =

Kasar

Kathi

6(a). Landed-dominant

West

Katkari-Kathodi

42(c). Hill tribe

Sahyadri

Kavar

6(b). Peasants

Cent. Prov.

Kayasth

4. Writers

Upper Ind. and Bengal

Kevat

14. Fishers etc.

Upjier India

Khaira =

Kora

Khalpo =

= Camar

Khambu

44. Himalayan tribe

N6pal

Khamti

43(g). Hill tribe

E. Assam

Khandait

6(a). Landed-dominant

Orissa

Khangar

19. Watchmen

Cent. Ind.

Kharia

42(a). Hill tribe

Bengal

Kharol

15. Salt-workers

Rajputana

Kharvar

42(a). Hill tribe

Bengal

Kharvi

15- Salt-workers

West

Khas

44. Himalayan tribe

Nepal

Khasf

43(C). Hill tribe

Assam

Khati

8(c). Carpenters

Upper India

Khatik

27. Butchers

Upper and West. India

Khatri

3. Traders

Panjab

Khatri

4. Writers

Ganges Valley

Khatri

9. Silk-weavers

West

Khavas

30. Domestic servants

West

Khojah

3. Traders

West

Khukar

6(a). Landed-dominant

Panjab

Khumra

34. Grindstone-makers

I'pper India

Kirar

6(b). Peasants

Cent. Prov.

Kisan

6(b). Peasants

Agra and Cent. India

Koc = Rajbansi

Kodagu

6(a). Landed-dominant

Coorg

Kodikkal

6(c). Bitel-vine-growers

Tamil

Koeri

6(b). Peasants

Agra, Oudh and Bihar

K6I

42(a). Hill tribe

Cent. Prov.

Kolhati =

Dombar

KoH

6(b). Peasants

West

Kolta

6(b). Peasants

Cent. Prov.

Komati

3. Traders

Telingana

Kondu-Dora

42(31 Hill tribe

N. E. Madras

Koraca =

Kuravan

Kora-Khaira

33. Earth-workers

Bengal

Korf

1

9. Weavers

Upper India

Korku-Korva

42(b). Hill tribe

Berar and Cent. Prov.

Korvf = Kuravan

i6o

5. Ethnography.

1
( asic r,roii|i

Locality

Kn^ti

9, Weavers

Uekkan and Cent. Prov.

K()ta

42(d). Hill tribe

Nilgiri

K<-)yi

42(a). Hill tribe

Cent. Prov. etc.

Ksatriya = Rajput

Kuki unsp''-

43(f). Hill tribes

Assam Frontier

Kumhar

II. Potters

Univ. except in South

Kunbi

6(b). Peasants

Dekkan and West

Kunjra

25. Greengrocers

Upper India

Kuravan-Koraca

36. Mat-makers

Telingana and Dekkan

Kuriccan

41. Fowlers

Malabar

Kurmi

6(b). Peasants

Upper India

Kurubar-Kurumban

32. Shepherds

South

Kurukh = Oraon

Kuruman

42(d). Hill tribe

Nilgiri

Kus'avan

II. Potters

Tamil

Kilta

30. Rice-pounders

Upper India

Labana

31. Carriers

Univ. except in East

Labbai

3. Traders

S. E. Coast

Lakhera

28. Lac-workers

Upper India

Lalung

43(a). Hill tribe

Assam

Lepca-Rong

44. Himalayan tribe

Sikkim

Lhota

43(ei. Hill tribe.

E. Assam

Limbu

44. Himalayan tribe

Nepal

Lihgayat unsp''-

6(b). Peasants

Karnatic

Lodha

6(b). Peasants

Upper India

Lohana

3. Traders

Sindh

Lobar

8(d). Blacksmiths

Univ. except in South

Luniya-Nuniya

15. Salt-workers

Upper India

Lusei

43(0- Hill tribe

E. Assam

Machi

14. Fishermen

Panjab and West

iVladiga

18. Leather-workers

Telingana

Mahant = Karan

Mahar

17. Field-labourers

Dekkan

McihilT

42(a). Hill tribe

Bengal

iMahtam

41. Fowlers etc.

Panjab

Majhvar

42(a). Hill tribe

S. Ganges Valley

IMal

19. Watchmen

Bengal

Mala

17. Field-labourers

Telingana

Malaiyan

42(d). Hill tribes

Nilgiri and Malabar

Male

42(a). Hill tribe

Bengal

Mair

6(c). Market-gardeners

Univ. except in South

Maliar

6(c). Market-gardeners

Panjab

Mallah unspJ-

14. Fishers and boatmen

Bengal

Appendix B. Caste Index.

i6i

Caste

Group

Malo

Mal-Paharia

Mang

Mahgala

Mangar

Manihar

Mappila

Maratha

Maravan

Marayan

Mayara

Mazbi

Mcc

j\lC-dar = Buruil

Mcgh

Meithci

INIcman

MC-0

Mihtar = Bhangi

Mikir

Mimar = Raj

Mma

Wirasi-Dum

Miri

Moci

Moger

Wohano

Mrung = Tipparah

IMughal

Mukkuvan

Munda

Murao

Murmi

Musahar

Mutraca

Naga unsp'i-

Nagarci

Nai-Nhavl

Naikada

Namas'udra

Nat

Nattaman

Nayak

Nayar

Indo-Aryan Research. II. I

14. Fishers and boatmen
42 (a). Hill tribe
18. Leather-workers
12. Barbers

44. Himalayan tribe
28. Bead-pedlars

3. Traders

6(a). Landed-dominant

6(a). Landed-dominant

12. Barbers etc.

26. Confectioners

20. Scavengers
43(a). Hill tribe

18. Leather-workers
43(f). Hill tribe

3. Traders
6(b). Peasants

43(d). Hill tribe

19. Watchmen

21. Genealogists
43(b). Hill tribe

18. Leather-workers
14. Fishermen
14. Fishermen

45. Muslim race
14. Fishermen
42(a). Hill tribe

6(ci. Market-gardeners
44. Himalayan tribe
17. Field-labourers

19. Watchmen

43 (cl Hill tribes

38. Drummers
12. Barbers
42(b). Hill tribe

6(b). Peasants

39. Acrobats
6(b). Peasants

42(b). Hill tribe
6(a). Landed-dominant

Locality

Bengal

Bengal

Dekkan

Telingana

Nepal

Upper India

Malabar

Dekkan etc.

Tamil

Malabar

Bengal

Panjab

Assam

Panjab Hills

Manipur

West

Rajputana and Panjab

Assam

Rajputana

Panjab

Assam

Univ. except in South

Kanara

Sindh

Upper and West. India

Malabar

Bengal etc.

Upper India

Nepal

Upper India and Bihar

Telingana

Assam

Upper India

Univ. except in South

West

Bengal

Upper India

Tamil

West

Malabar

l62

5. Ethnoorachy.

Caste

fir(iU|i

Locality

Nevar

44. Himalayan tribe

Nipal

Neyige unsp"*-

9. Weavers

Karnatic

Nhavl = Nai

Nihal

42(b). Hill tribe.

West

Nllari

29(b). Indigo-dyers .

Upper India

Niyariya-Dhuldhoya

8(b). Gold-dust-washers

Upper and West. India

Nora

43 (g). Hill tribe

E. Assam

Nuniya = Luniya

Od-Vatldar

33. Earth-workers

Univ. except in East

Oraon-Kurukh

42(a). Hill tribe

Bengal

Osval

3. Traders

West

Paik

16. Toddy-drawers

Kanara

PaUan

17. Field-labourers

Tamil

Palle

14. Fishermen

Telingana

PaUi

17. Field-labourers

Tamil

Panan

22. Exorcists

Malabar

Paficala

8 (a). Artisans

Karnatic

Paficamasale

6(b). Peasants

Karnatic

Pandaram

23 (a~l. Priests

Tamil

Panisavan

5. Devotees

Tamil

Panka-Pan

9. Weavers

Cent. Prov.

Paraiyan

17. Village servants

Tamil

PardhT

41. Fowlers etc.

Dekkan

Pant = DhobI

Parivaram

30. Domestic servants

Tamil

Pasi

16. Toddy-drawers

Upper India and Bihar

Patella

42(b). Hill tribe

West Belt

Pathan

45. Muslim race

N. W. Frontier

Pathari

42(a). Hill-tribal-priests

Cent. Prov. etc.

Patharvat

15. Stone-w^orkers

Dekkan

Patnl

14. Fishers etc.

Bengal

Patra-Pator

28. Pedlars

Orissa

PattunOrkaran

9. Silk-weavers

Tamil

Patve

9. Silk-weavers

Upper and Central India

Pendhari

31. Carriers

Dekkan and Karnatic

Perike

9. Hemp-weavers

Tamil

Phakial

43(g). Hill tribe

E. Assam

Phulari-Hugar

23(b). Temple servants

Dekkan etc.

Pinjari

29(c). Cotton-scutchers

West

Pod

6(b). Peasants

Bengal

Poroja

42(a). Hill tribe

N. E. Madras

Porval

3. Traders

Rajputana etc.

Prabhu

4. Writers

West

Appendix B. Caste Index.

I6J

Caste

Group

Locality

Pradhan

42(a). Hill tribe

1
Cent. Prov.

Pujarl

23 (a). Hill-tribal-priests

Panjab Hills

Pulayan-Ccruman

17. Field-labourers

Malabar

Qasab

27. Butchers

Upper India

Quresi = Sckh

Rabarl

7. Camel-breeders

Rajputana etc.

Rabhfi

43iat- Hill tribe

Assam

Raj-Mimar

8(c). Masons etc.

Upper India

Rajbaiisl-Koc

6(a). Landed-dominant

Assam and Bengal

Raj-Bhat

21. Bards and genealogists

Bengal

Rajput-Ksatriya

2. Landed-dominant

Upi)er and West. India

Rajvar

17. Field-labourers

Bengal

Ramaiya

28. Pedlars

Panjab

RamOs'i"

19. Watchmen

Dekkan

Rangrej

29 (b\ Dyers

Univ. except in South

Rathi

6(bi. Peasants

Panjab Hills

Raut

6(b). Peasants

Panjab Hills

Razu

6(a). Landed-dominant

Telingana

Redc.li = Kapu

Rehgar

15. Salt-workers

Rajputana

Rengma

43 (ei. Hill tribe

E. Assam

Rong = Lepca

Sadgop

6(bV Peasants

Bengal

Sadhu unsp*^-

5. Devotees

West

Sagirdpesa

30. Domestic servants

Orissa

S'aha = Suhri

Sahariya

41. Fowlers etc.

Cent. India

Saint

6(c). Market-gardeners

Panjab

Sainteng

43(c). Hill tribe

Assam

Saiyad

45. Muslim race

Universal

S'akkiliyan

18. Leather-workers

Tamil

Sale

9. Weavers

Dekkan and South

Sammo

6(a). Landed-dominant

Sindh

Samru

6(a). Landed-dominant

Sindh

San

43(g)- Hitl race

E. Assam

S'anan

16. Toddy-drawers

Tamil

S'aiikhari

28. Armlet-makers

Bengal

Sahsiya

40. Thieves

Panjab

Santal

42(a). Hill tribe

Bengal

Satani

23(b). Temple servants

Telingana

Savara

42(a). Hill tribe

S. Orissa

Segidi

16. Toddy-drawers

Orissa

1 64

S. Ethnography.

Caste

( I r . I u 1 1

Locality

Sckh-QurC-.-ji

45. Muslim race

Bengal

Sema-Sima

43(e). Hill tribe

E. Assam

S'embadavan

14. Fishermen

Tamil

Scnaikku(laiyan

61c). Bitel-vine-growers

Tamil

Sevak

23. Priests to Jains .

Rajputana

S'ikligar

34. Knife-grinders

Upper and West. India

Sima = Sema

S'impT

29(a). Tailors

Dekkan

Singpho

43(h). Hill tribe

Assam

Sonar

8(b). Goldsmiths

Univ. except in South

S'rimair

3. Traders

West

Subarnabanik

3. Traders

Bengal

S'udra

30. Domestic servants

Bengal

Sumro

6(a). Landed-dominant

Sindh

Sunrl-S'aha

29(d). Distillers

Bengal

Sunuvar

44. Himalayan tribe

Nepal

Sutar

8(c). Carpenters

Univ. except in South

Tadvi

42(b). Hill tribe

West

Taga

6(a). Landed-dominant

Agra

Takari-Takankar

34. Grindstone-makers

Dekkan

Talavia = Dubia

Tambala

23(a). Priests

Telingana

Tambat

S(f). Coppersmiths

West

Tamboli

25. Bitel-sellers

Univ. except in South

Tandan

16. Toddy-drawers

Malabar

Tanti

9. Weavers

Bengal

Tantva

9. Weavers

Bihar

Tarkhan

8(c). Carpenters

Panjab

Telaga

6(b). Peasants

Telingana

Tclf-Ghancr

10. Oil-pressers

Univ. except in South

Tengima = Angami

Thakar

6(b). Peasants

Panjab Hills

Thathera

8(f). Brass-workers

Upper India

Thavi

8(c). I^Lisons

Panjab Hills

ThorT

31. Carriers

Panjab Hills

Tigaja

6(c). Market-gardeners

S. Dekkan

Tipparah-Mrung

43(a). Hill tribe

E. Bengal

Tiyan

16. Toddy-drawers etc.

Malabar

Tiyar

14. Fisher and boatmen

Bengal

Toda

42(d). Hill tribe

Nilgiri

Togata

9. Weavers

Karnatic

Tottiyan

6(b). Peasants

Karnatic

Turaha

38. Drummers etc.

Bengal

Turl

35. Bamboo-workers

Bengal

Appendix B. Caste Index.

165

Caste

Ci r 0 u p

Locality

Turk unsp''-

45. Muslim race

Panjab West

Turung

43(g). Hill tribe

E. Assam

Ulama

23(a). Priests

Panjab

Uppara

15- Salt-workers

Karnatic

Uppiliyan

15. Salt-workers

Malabar

Vaddar = Oil

Vaduga

3. Traders

Telingana

Vaghri

41- Fowlers

West

Vaidya

4. Writers

Bengal

Vakkaliga

6(a). Peasants

Karnatic

Valaiyan

41. Hunters

Tamil

Valjuvan

23(a). Low priests

Tamil

Vaniyan

10. Oil-pressers

Tamil

Vannan

13. Washermen

Tamil

Varlf

421CI. Hill tribe

Sahyadri

Vedan

41. Hunters

Tamil

Velama

6iai. Landed-dominant

Telingana

Velan

22. Exorcists

Malabar

Veimian

6(b). Peasants

Tamil

Veluttedan

13. Washermen

Malabar

Vettuvan

41. Hunters

Tamil

Vidhur

4. Writers

Dekkan and Cent. Prov.

Vohora = Bohra

Yakha

44. Himalayan tribe

Nepal

Yanadi

42(d). Hill tribe

Telingana

Yata

16. Toddy-drawers

Orissa

Yerukala

36. Mat-makers

Telingana

5. Ethnography.

APPEND

Showing (A) the number returning each principal '.

of the populatio

Language and Family

India

No. per

10,000

of popu-
language I ,atio„

Total number
returning the

N. West

II.

I. Kol-Khervari
Kol

Santali
Savara
Kharia
Korku .
Gadaba
Kora .
0//iers .
Dravidian
Gurx.l .
Oraon .
Kand .
Maltc) .
Telugu
Kanarese
Kodagu
Tulu .
Tamil .
Malayalam
Brahiii
0//urs

III. Gipsy tongues

IV. Indo-Aryan
Sina etc

I Kasmiri
I Lahnda
[ Sindhi .
[ West Pahari
I Central Pahari
I East Pahari
[ West Hindi
I Panjabi .
I Rajasthani
1 Gujarati .
East Hindi

>) Including N>ii

3,179,300
948,700
1,790,500
157,100
102,000
87,700
37.200
23,900
32,200

56,315.700

1,125,500

592,300

494,100

60,800

20,600,000

10,364,700

39,200

535.200

16,425,000

6,028,900

47,900

2,100

344,100

219,352,100

54,200

1,007,900

3,337,900

3,002,800

1,710,000

1,270,900

138,300

40,568,900

17,033,300

10,917,100

9,921,700

22,136,400

112

33
63
6

4
3
I

40
21
17

728
366

I

19
581
213

2

1

12
7.756

2

36
118
106
60
45
5
1.434
602
386
351
783

-')

-

I

0 1

l

0

I

0

0

0

—

9

4

9.380

9494

193

0

3.550

3

0

1,244

0

10

552

579

—

0

3

4

6

1.559

4.624

5,833

452

245

0

0

—

—

Stales

ecicd with the Pr

I Including the N. \V, Frontier Provi

Table of Languages.

3LE I.

^age, and (B) the Linguistic distribution per 10,000
eh Province or State.

B

. 1 entral

Central

East

West

South

a
c >^

a 0

'■5 ><

■(3

E

u

>>

It

■a

Xi

at

a.

3 0

c E

bo

as

u

Xi

0

u

V >

u 2

(X,

(U

<

CC

E

0

OQ

10

—

0

72

354

121

103

46

0

io

—

—

16

112

61

—

—

—

—

^

—

220

49

—

—

—

37

0

;-

—

—

6

10

3

—

—

—

—

—

—

!—

—

—

50

—

—

103

—

—

—

—

—

J—

—

—

0

—

—

—

—

—

—

9

—

i-

—

—

—

3

0

—

—

—

—

—

—

-

—

—

0

9

8

—

—

—

—

—

—

,0

0

27

934

86

51

437

1,448

2

6,122

9.191

9,260

}-

0

25

751

—

3

302

1

—

68

12

—

;-

—

—

41

69

17

—

—

—

—

—

—

p

—

—

45

7
8

19

—

—

—

—

88

—

P

0

I

89

8

129

49

2

4,621

3,381

1,507

,D

—

0

3

0

0

4

1,354

0

1,402

372

7.301

l~

—

—

—

—

—

—

—

—

—

0

0

\-

—

—

—

—

0

—

0

—

0

118

37

D

0

I

5

0

4

2

4

0

31

3.805

409

-

—

—

—

0

—

0

0

—

0

1,415

6

I

_

_

.

0

0

3

2

10

21

0

0

250

13

I

109

8

69

»89

9.996

9.956

8,962

9,494

7,688

9.295

8,510

9,554

3.747

743

645

3

0

_

0

0

—

0

0

0

0

—

D

0

0

—

—

—

—

—

—

—

—

—

D

40

I

0

0

0

—

90

I

0

0

0

D

—

—

0

—

—

—

—

—

—

—

—

II

0

—

0

—

—

0

—

—

0

0

—

5

0

0

—

ID

32

—

0

—

—

0

—

527

2,823

5,479

1,629

171

29

991

495

354

1,069

212

48 >

3

23

4

I

0

5

2

I

0

2

0

0

2

6,743

2,171

430

I

12

152

46

10

54

0

2

I

360

326

17

I

2

76

3,228

9,431

I8

23

6

.25

—

1,623

3,653

146

545

15

I

0

0

0

—

lilank

means that

the langu

ige was no

returned,

a cipher tl

a. it was r

etuniedby

less than

ane in io,oe

xtoflbepo

pulation.

I68

5. ExHNOGRAPm-.

India

N. West

Language and Family

No. per

10,000

of popu-
anguage ^^^^^^

Total number
returning the

Hihari .
Bengali
Assamese
Oriya .
Marathf
0//iers .
Iranian
Pasta

VL

Baluci

Persian

Of/urs

Tibeto-Burman . .

Bhotia

Kanavari

Kiranti

Murmi

Of/ier Himalayan . .

Miri-Abor

0//u-r East-Himalayan

B(-)do

Garo

Tipparah

0//ier Assam. . . .

Mikir

Naga languages . .

Meithci

Lu.sC'i ....

VII.

VIII

IX.

X.

XI.

XII

XI IL

Total

Kuki

Of/t£rs

Kacin

Burmese

Mni . .

Tai (Sam

Mon (Khasi)

Mongolian

Malay

Semitic (Arabic). . .

Hamitic

European

Knglish

Of/nrs ' '. '

i'nspicifted^\

population returning language

■) Kelurncd by Icsi than one per 10000

34,579,800

44,413,600

1,350,800

9,674,200

18,233,200

800

1,388,200

1,218,500

150,600

18,900

200

1,804,800

244,900

19,500

45.400

32,200

83,800

40,800

900

239,500

185,500

112,000

59,000

83,600

164,160

269,300

72,200

53,900

20,000

1,800

65,400

10,500

3.400

177,800

3.600

26

19,700

180

243,100

227,900

15,200

282,832,000

the Province or 11

1.223

1,570

48

342

645

O

49

43

5

I

o
64

8

I

2

597
597

Table of Languages.

169

c

tral

Central |

East I

West 1

South

a >-.

•a 0

IS

'■5 >>

. V

v<

2 ^

SI

c

03

5
a

<

u
a
u
u
CQ

0

B

CQ

•s

2

n

03

•a

u

(«

X

<n

2

•0

0

0

00

)

I

3
0
0

2

0
0

0
0
0

—
0

0

7
7
0
2

3
0

2

I
I

0
0

0

0

I
I
0

222
0

0
130

I

I
0
0

0

5
5
0

I

0

I

1,355
1,876

0
0

0

0

0

0

0
0
6
6
0
5

3,095
5,279

0
790

0

I

0

0

0

0

59

3

5
4
10

2
5

13
3

2

2
I

8

I

0

0
0
6
6
0
I

I
4,812
2,203

38

9

3

2

I
0

1,835
2

I

0

9
66

I
388
217

17
96

134
268

417
"7
76

22

3
I

5
289

0

0
0
3
3
0

5

0
0

7,969

2

2
0

0

0

0

2
2
0

I

0
0
0
0
4,649

5
0

3

2

0

0

2
0
20
•5

5

2

0
0

198

0
0
0
0

0

I
I

0

2

0
0

2
2,602

I
I

0

0
0

9

7
7
0

4

0

433
7S
0
0
0
0
0

0

0

0
0
0

10

9
I

3

0

1

IS"
0

4
0

4

0

0
0
0

18

18
0
4

170

5. Ethnography.

APPENDIX, TABLE 11.

Religions per 10,000 of population of each division.

ll

Brahmanic

»

V

c
.2

Political

'3

0
B

.c

.,

j=

'C3

£

u

V

Division

h|

c

u
03

a

•0
•0

a

Oh

""*

'u

U

0

2 / Kashmir . .

2,372

89

I

121

7,416

I

^1 Panjab* . .

—

3,898

—

792

19

3

—

5,261

—

27

—

2; ' Sindh* . .

—

2,309

—

—

3

—

6

7,652

I

23

6

__ I Rajputana

371

8,320

I

2

352

—

—

951

—

3

—

2 Unit. Prov.*

—

8,532

14

3

17

—

I

1,412

—

22

—

S 1 Central India

1,150

8,094

—

2

131

—

—

613

—

9

1

^ 1 Centr. Prov. *

1,469

8,208

—

I

41

—

I

259

—

21

—

Bombay*. .

43

8,689

—

—

243

—

35

889

5

96

—

w Baroda . .

903

7,922

—

—

248

—

43

84s

—

39

—

^ Berar .

472

8,671

—

5

71

—

2

770

—

9

—

Haidarabad .

59

8,860

—

4

18

—

I

1037

—

21

—

1 ( Bengal* . .

354

6,330

—

—

I

30

—

3,248

—

36

I

W \ Assam* . .

1,744

5,597

I

I

3

—

14

2,581

—

59

—

1 ( Madras* . .

166

8,916

7

642

269

o\ Mysore . .

156

9,205

—

—

25

—

—

523

—

91

—

India . .

289

7,305

4

77

47

10

3

2,167

—

98

— •

Including Native States.

Tables of Religion.

171

I

APPENDIX, TABLE III.

Showing the numerical strength of the principal Forest Tribes, and the
relative prevalence of the Tribal language and religion.

Total

Per-
cciuage

I'c

centage returning Tribal Religion

1 Tribe

popula-

retur-
lan-

tion

Total

Provincial

A. Central Belt

9.178,515*

58

Santal . .

1,907,871

94

67

Bengal 70; Assam (labourers) 7

Munda

466,668

J62

65]
87 71

Bengal 78; Assam (labourers) 7

,_

Ho .

385,125

Bengal

>

K61 .

298,997

1

56)

Cent. Prov. 22; C.lnd. lOO; Elsewhereo

U

Korku

151755

48

45

Berar 94; Cent. Prov. 13

wC

«

Savara

367,367

43

45

Madras 87; Cent. Prov. 5; Bengal 0

:i

Kharvar

•39,625

—

1

Bengal \ ; Cent. Prov. 10

^

Kharia

120,725

92

55

Bengal 69; Cent. Prov. 47

Khaira

109,571

—

13

Bengal 6; Cent. Prov. 47

Bhinjia

84,990

—

31

Bengal 0; Cent. Prov. 33

Gond.

2,286,913

45

72

Berar92;C.Prov.77; Beng.27;Madras3

Gorirhr

264,605

68

Cent.Ind. 100; Un. Prov. 0; Bengal 0

.2

KoyI .

115,216

10

Madras 17; Haidarabad I

|5

Poroja

91,886

29

Madras

a

Pan .

684,746

—

4

Bengal 6; Cent. Prov. and Madras I

Q

Oraon

614,501

96

71

Bengal 73; Assam (labourer) 8

Kand

701,198

70

68

Madras 82; Cent. Prov. 57; Bengal 38

Ot/iers

3Sb,S4b

-

48

B. W^estern Belt

2,175.514

—

45

Bhil . . .

1,198,843

64

55

Cent.Ind. lOO; Baroda 100; Rajput. 97;
Bombay 14; Berar 57

Bhilala . .

144,423

—

91

Cent. Ind. 100

Kotval

53,342

58

Cent. Ind. 100

_

Tadvi

10,566

80

Bombay (rest Muslim)

'-2

Dhodia

110,242

—

17

Baroda 100; Bombay 3

Dubla

129,267

—

24

Baroda lOO; Bombay 3

Naikada

115,600

—

lO

Cent. Ind. 100; Bombay 8

Varll .

152,309

—

—

Bombay

KatkarF

93.032

—

2

Bombay

Others

ib§,88i

—

43

Not including Christian Converts.

172

5- Ethnography.

; Total

Pcr-

Percentage returning Tribal Religion

Tribe

popula-

ning
Tribal

tion

lan-
guage

Total

Provincial

C. Nllgiri

302,392

—

9

Irula .

86,087

2

—

Kuruman

179,928

5

13

Toda .

807

99

99

Kota .

1271

—

45

Badaga

34,299

98

—

D. North-East

1,419,222

—

76

Kacarf . .

242,904

1-

71

Assam 71 ; Bengal 79

Mcc .

99,534

78

Assam 100; Bengal 15

Tipparah

111,279

lOI*

4

Assam 49; Bengal 0

Garo .

166,237

112*

95

Assam 99; Bengal 82

Rabha.

67,285

30

89

Assam

Lalung

25,513

46

100

Assam

Naga .

162,797

—

99

Assam

Mikir .

87,335

96

99

Assam

Kuki .

67,212

—

86

Assam 100 ; Bengal 0

Lusei .

63,588

113*

100

Assam

Miri .

46,720

87

49

Assam

Cutiya

85,829

3

0

Assam

KhasF .

159,549

III*

99

Assam

Others.

59fi53

"

97

Assam

* The Tribal language is here returned by some no longer returning the Tribe.

\

Indl\, general. 173

A LIST

of the more important works on Indian Ethnography

by Dr. W. Siegling.

India, general.

The Census of India, 1901, Vols. I— XXVI.

(Vol. I: India; Part I: Report by H. H. Risley and E. A. Gait; Part II:
Tables; Part III: Ethnographic appendices. 3 vols. fol. Calcutta 1903).

The Imperial Gazetteer of India, by W. W. Hunter. 9 vols. London 1881 ; —
2"'' ed. 14 vols. London 1885 7; — 3'' ed. 26 vols. Oxford I907'9.

The Imperial Gazetteer of India, provincial series. 1907 sqq.

Farla y Sousa, M. de. Asia Portuguesa ... 3 vols. fol. Lisboa 1666—75; —
The Portugues Asia; or, the history of the discovery and conquest of
India by the Portuguese; translated by J. Stevens. 3 vols. London 1695.

Lettres edifiantes et curieuses, ecrites des missions ctrangcres. En 34
recueils. 12° Paris 1717 — 74; — 2= ed. 26 tomes 12° Paris 1780 — 83;
— autre ed. 26 tomes 12° Toulouse 1 8 10; — 4 tomes 8° Paris

1837—43; --
Nouvelles lettres edifiantes des missions de la Chine et des Indes Orien-

tales. 8 tomes 120 Paris 1S18 — 23; —
Ed. nouvelle: Lettres edifiantes et curieuses, ecrites par des mission-
naires de la Compagnie de Jesus; collationnces sur les mcilleures
editions, et enrichies de nouvelles notes. Mem. du Levant, t. 1—9;
mem. d'Amerique, t. 10—16; mem. des Indes, t. 17 — 24; mem. de la
Chine, t. 25 — 38; mem. des Indes et de la Chine, t. 39 — 40. 20 vols.
8" Paris 1829—32.

Hamilton, Capt. A. A new account of the East Indies: being the obser-
vations and remarks of Capt. A. Hamilton, who resided in those parts
from the year 16SS to 1723. 2 vols. Edinburgh 1727; — 2'"' ed.
2 vols. London 1739; — 3'' ed. 2 vols. London 1744.

TiEFFENTHALER, J. Historisch-geographische Beschreibung von Hindustan.
Aus dessen latein. Handschrift iibersetzt. Herausgeg. von J. Bermmlli.
2 Bde. Berlin und Gotha 1785—86; — Description historique et
geographique de I'lnde, qui prescnte en 3 vols., enrichis de 68 cartes
et autres planches: i) La geographic de I'lndoustan, ccrite en latin,
dans le pays meme, par le pere Joseph Tieffenthaler. 21 Des recherches
historiques et chronologiques sur I'Inde, et la description du cours
du Gange et du Gagra, avec une tres grande carte, par Anquetil
du Perron. 3) La carte gencralc de I'Inde, celles du cours du Brahma-
poutra, et de la navigation interieure du Bengale, avec des memoires
relatifs a ces cartes, publics en anglois, par Jacques Rennell. Le tout,
augmente de remarques et d'autres additions, redigd et publid en
franqois, par Jean Bernoulli. 3 vols. 4° Berlin 1786—91.

Rennell, Maj. J. Memoir of a map of Hindoostan, or, the Mogul empire ;
and a map of the countries between the Indian rivers and the Caspian,
account of the Ganges and Barrampooter rivers, etc. 4° London 1788;

174 5- Ethnography.

— 2"'' ed., with additions, corrections, etc. 4" London 1792; —
S"* ed., with additional map and geography of the peninsula of India.
4" London 1793; —
(Trad, frangaise, par J. B. Boucheseiche, etc. 3 tomes Paris 1800).

Forbes, J. Oriental memoirs; written during 17 years' residence in India
(1766 — 84), including observations on parts of Africa and South America,
and a narrative of occurrences in four Indian' voyages. 4 vols. 4°
London 1813; — 2"'^ ed., abridged, 2 vols. 8° and i vol. 4" (illustra-
tions) London 1834 5.

Hamilton, W. The East India gazetteer; containing particular descriptions
of the empires of Hindostan and the adjacent countries; India beyond
the Ganges, and the Eastern Archipelago; together with sketches ... of
their various inhabitants. London 1815; — 2"'' ed. 2 vols. London 182S.

Craufurd, Q. Researches concerning the laws, theology, learning, com-
merce, etc., of ancient and modern India. 2 vols. London 1 81 7.

Hamilton, W. Geographical, statistical, and historical description of Hin-
dostan and the adjacent countries. 2 vols. 4" London 1820.

Langles, L. Monuments anciens et modernes de I'Hindostan, dccrits sous
le double rapport archeologique et pittoresque, et precedes d"une
notice geographique, d'une notice historique, et d'un discours sur la
religion, la legislation et les moeurs des Hindous. 2 vols. fol. Paris 1821.

Remusat, a. Melanges asiatiques, ou choix de morceaux critiques et de
memoires relatifs aux religions, aux sciences, aux coutumes, a I'histoire
et a la geographic des nations orientales. 2 tomes Paris 1S25/6; —
Nouveaux melanges asiatiques. 2 tomes Paris 1829; —
Melanges posthumes. Paris 1843.

Colebrooke, H. T. Miscellaneous essays. 2 vols. London 1837.

Hough, Rev. J. History of Christianity in India, from the commencement
of the christian era. 5 vols. London 1839.

Benfey, Th. Indien. 4° Leipzig 1840 (Part of "Ersch und Gruber's
Encyclopaedic").

Parkes, Mrs. F. Wanderings of a pilgrim, in search of the picturesque,
during the four and twenty years in the East; with revelations of life
in the Zenana. 2 vols. London 1850.

Balfour, E. Cyclopaedia of India and of Eastern and Southern Asia.
Madras 1857—62; — 2"'' ed. 5 vols. Madras 1871 — 73; — 3'' ed.
3 vols. London 1885.

Lassen, Ch. Indische Altertumskunde. 4 Bde. (i and 2 : 2. Aufl.) Leipzig
1858-74.

Martin, R. M. The Indian empire : its history-, topography, government,
finance, commerce, and staple products. With a full account of the
native troops, and an exposition of the social and religious state of
one hundred million subjects of the crown of England. 3 vols.
London 1858—61.

Prinsep, J. Essays on Indian antiquities, historic, numismatic, and pala;o-
graphic; to which are added his useful tables, illustrative of Indian
history, chronology, modern coinages, weights, measures, etc. Ed. by
Edw. Thomas. 2 vols. London 1858.

Malleson, G. B. An historical sketch of the native states of India in sub-
sidiary alliance with the British government. With a notice of the
mediatized and minor states. London 1875.

India, general. 175

Tagore, S. M. Yantra Kosha, or, a treasury of the musical instruments
of ancient and of modern India, and of various other countries.
Calcutta 1875.

Wheeler, J. T. Early records of British India; a history of the English
settlements in India, as told in the government records, the works of
old travellers, and other contemporary documents, from the earliest
period to the rise of the British power in India. Calcutta 1878; —
2"'' ed. 1879.

Egerton (of Tatton, Lord) W. An illustrated handbook of Indian arms ;
being a classified and descriptive catalogue of the arms exhibited at
the India Museum, with an introductory sketch of the military history
of India. London 1880: — new ed., with considerable additions:
A description of Indian and oriental armour. Illustrated from the
collection formerly in the India Office, now exhibited at South Ken-
sington, and the author's private collection. London 1896.

Hodgson, B. H. Miscellaneous essays relating to Indian subjects. 2 vols.
London 1880.

Phear, Sir J. B. The Aryan village in India and Ceylon. London 1880.

Schlagintweit, E. Indien in Wort und Bild. Eine Schilderung des indischen
Kaiserreiches. 2 Bde. fol. Leipzig i88o'i; — 2. Aufl. 1890/1.

Hunter, W. W. The Indian empire : its peoples, history, and products.
London 1882; — 2"<' ed. 1886; — 3'' ed. 1892.

Ly.\ll, Sir A. C. Asiatic studies, religious and social. London 1882: —
2"'! ed. 1S84; — First and second series. 2 vols. London 1899; —
2"'! ed. 1907.

LiTH, P. A. VAN der. Livre des merveilles de I'lnde. Texte arabe. Tra-
duction franqaise par L. Marcel Devic. 4° Leide 1883—86.

Whitworth, G. C. An Anglo-Indian dictionary ; a glossary of Indian terms
used in English, and of such English or other non-Indian terms as
have obtained special meanings in India. London 1885.

Statistical atlas of India, fol. Calcutta 1886; — 2'-'i ed. fol. Calcutta 1895.

Yule, H., and A. C. Burnell. Hobson-Jobson : being a glossary of collo-
quial .Vnglo-Indian words and phrases of kindred terms, etymological,
historical, geographical, and discursive. London 1886; — 2"'' ed.,
revised by W. Crooke. London 1903.

Le Bon, G. Les civilisations de llnde. 4° Paris 1887; — nouv. ed.,
augmentce. fol. Paris 1900.

Bastlan, a. Ideale Welten in Wort und Bild. 3 Bde. 4" Berlin 1892.

BosE, P. N. History of Hindu civilisation during British rule. 3 vols.
Calcutta 1894—6.

Chakrabarti, J. Ch.\sdra. The native states of India. Calcutta 1895; —
2"<' ed. London 1896; — S"* ed. London 1897.

Baden-Powell, B. H, The Indian village community. Examined with refe-
rence to the physical, ethnographic, and historical conditions of the
provinces; chiefly on the basis of the revenue-settlement records and
district manuals. London 1896.

The Hind Rajasthan, or, the annals of the native states of India. Com-
piled by Markand Nandshankar Mehta and Manu Nandshankar Mehta.
Baroda 1896.

176 5- Ethnography.

Temple, G. Glossary of Indian terms, relating to religion, customs, govern-
ment, land, and other terms and words in common use. London 1897.

Baden-Powell, B. H. The origin and growth of village communities in
India. London 1899; — new ed. London 1908.

La Mazeliere, Marquis de. Essai sur revolution de la civilisation indienne :
rinde ancienne; I'lnde au moyen age; I'lndc moderne. 2 tomes
Paris 1903.

Linguistic survey of India. Ed. by G. A. Grierson. fol. Calcutta 1903 ff.

Crooke, W. Things Indian. Being discursive notes on various subjects
connected with India. London 1906.

Manlxci, N. Storia do Mogor; or, Mogul India, 1653— 1708. Translated,
with introduction and notes, by W. Irvine. 4 vols. London 1907—8!

Ancient India.

Robertson, W. An historical disquisition concerning the knowledge which
the ancients had of India; and the progress of trade with that country
prior to the discovery of the passage to it by the Cape of Good Hope.
With an appendix, containing observations on the civil policy, the
laws and judicial proceedings, the arts, the sciences, and religious
institutions of the Indians. 4° London 1791; — (8° Basil 1792;
Utrecht 1792); — 3'^ ed. London 1799; — 7th ed. London 1817; —
12° Edinburgh 1806; — London 1821; — reprinted Calcutta 1904.
(Historische Untersuchung iiber die Kenntnisse der Alten von Indien.
Nebst Anhang, welcher Bemerkungen iiber die gesellschaftiichen Ver-
haltnisse usw. der Indier enthalt. Ubersetzt von G. Forster. Berlin 1792;
Recherches historiques sur la connaissance que les Anciens avaient de
rinde. Paris 1792.)

Reinaud. Fragments arabes et persans incdits, relatives a I'lnde, ante-
rieurement au ii« siccle de I'cre chretienne, recueillis avec traduction
frangaise. Paris 1845.

Reinaud. Memoire gcographiquc, historique et scientifique sur I'lnde,
antcrieurement au milieu du XI<= siccle de I'^re chretienne, d'apres
les ccrivains arabes, persans et chinois. 4" Paris 1849.

Speir, Mrs. Life in ancient India. London 1856.

Vivien de St. Martin, L. Etudes sur la gcographie grecque et latine de
rinde, et en particulier sur I'lnde de Ptolcmee, dans ses rapports
avec la geographic sanscrite. 4° Paris 1S5S.

Vivien de Pt. Martin, L. Etude sur la gcographie et les populations pri-
mitives du nord-ouest de I'lnde d'apres les hymnes vcdiques. Paris i860.

Manning, Mrs. (Formerly Mrs. Speir). Ancient and media:val India, being
the history, religion, laws, caste, manners and customs, language,
literature, poetry, philosophy, astronomy, algebra, medicine, archi-
tecture, manufacture, commerce, etc., of the Hindus, taken from their
writings. 2 vols. London 1869.

Mc Crindle, J. W. Ancient India as described by Megasthenes and Arrian.
Bombay 1S77.

ZiMMER, H. Altindisches Leben. Die Cultur der vedischen Arier, dargcstellt
nach den Samhitas. Berlin 1879.

Ancient India. — Travels. 177

Rajendralala Mitra. Indo-Aryans: contributions towards the elucidation
of their ancient and medieval history. 2 vols. London 1881.

Mc Crindle, J. W. Ancient India as described by Ktesias the Knidian.
Bombay 1882.

Mc Crindle, J. W. Ancient India as described by Ptolemy. Bombay 1885.

Alberuni. India. An account of the religion, philosophy, literature, geo-
graphy, chronology, astronomy, customs, laws, and astrology of India,
about A. D. 1030. Ed. by E. Sachau. 40 London 1887. — English trans-
lation with notes and indices by Ed. Sachau. 2 vols. London 1888;
— new ed. 2 vols. London igio.

Dl'tt, R. Ch. History of civilisation in ancient India, based on Sanscrit
literature. 3 vols. Calcutta 1889/90; — revised ed. 2 vols. London 1893.

Caland, W. Altindischer Ahncncult. Leiden 1893.

Mc Crindle, J. W. The invasion of India by Alexander the Great, as des-
cribed by Arrian, Q. Curtius, Diodorus, Plutarch, Justin, and other
classical authors. l^ondon-Westminster 1893: — 2"'' ed. London 1896.

Calanu, \V. Die altindischen Toten- und Bestattungsgebrauche. Amster-
dam 1896.

FiCK, R. Die sociale Gliederung im nordostlichen Indien zu Buddhas Zeit.
Kiel 1897.

Goblet d'Alviella. Ce que I'lnde doit a la Grece. Des influences classiques
dans la civilisation de I'lndc. Paris 1897.

Hillebrandt, a. Alt-Indien. Kulturgeschichtliche Skizzen. Breslau 1899.

Mc Crindle, J. W. Ancient India as described in classical literature: a
collection of Greek and Latin texts from Herodotus and other works,
translated and annotated. London 1901.

Smith, Vincent A. Asoka, the Buddhist emperor of India. Oxford
(London) 1901.

Hardy, E. Konig Asoka; Indiens Kultur in der Blutezeit des Buddhismus.
Mainz 1902.

Ray, Pr. Ch. History of Hindu chemistry from the earliest times to the
middle of the le'^ century A. D. Calcutta 1902 : — 2"<i ed., revised
and enlarged. London 1907.

Henry, V. La magie dans I'lnde antique. Paris 1903; — 2' ed. Paris 1909.

Rhys Davids, J. W. Buddhist India. London 1903.

S.WTH, V. A. The early history of India. Oxford 1904; — 2"'' ed. 1908.

LiJDERS, H. Das Wiirfelspiel im alten Indien. Berlin 1907.

V.\iDYA, C. V. Epic India, or, India as described in the Mahabharata and
the Ramayana. Bombay 1907.

Caland, \V. Altindische Zauberei. Darstellung der altindischen Wunsch-
opfer. Amsterdam 1908.

Travels.

Fa-Hien: Foe Koue Ki, ou relation des royaumes bouddhiques: voyage
dans la Tartaric, dans I'Afghanistan et dans I'lnde, execute a la fin
du IV<^ siecle, par Chy Fa Hian. Traduit du chinois et commente
par Abel Remusat. Ouvrage posthume, revu, complete, et augmentd
d'eclaircissements nouveaux par Klaproth et Landresse. 4° Paris 1836;
— (English translation of this ed. Calcutta 1848); —

Indo-Aryan Research. 11.5. "

17^ 5- Ethnography.

Travels of Fah-Hian and Sung-Yun, Buddhist pilgrims, from China to
India, 400 A. D. and 518 A. D. Translated from the Chinese by S. Beal.
London 1869; —

Record of the Buddhistic Kingdoms (by Fa-IIiam. Translated from the
Chinese by H. A. Giles. London i88o; —

Fa-Hien, a record of Buddhistic Kingdoms; being an account of his travels
in India and Ceylon (399—4141. Translated and annotated by J. Legge.
4" Oxford 1886.
Hiuen-Thsang. Histoire de la vie de Hiouen-Thsang, et de ses voyages
dans I'lnde, depuis I'an 629 jusqu'cn 645, par Hoeili et Yen-Thsong,
suivie de documents et d'cclaircissements geographiques tires de la
relation originale de Hiouen-Thsang. Traduit du chinois par Stanislas
Julien. Paris 1853; —

Mc-moires sur les contrces occidentals, traduits du Sanscrit en chinois,
en I'an 648, par Hiouen-Thsang, et du chinois en frangais par Stanislas
Julien. 2 tomes Paris 1857 '8; —

Hiuen-Tsiang, Si-yu-ki, Buddhist records of the western world. Trans-
lated from the Chinese by S. Beal. 3 vols. London 1884 88; — new ed.
2 vols. London 1906; —

Walters, Th. : On Yuan Chwang's travels in India, 629 — 45. Ed. after
the author's death by T. W. Rhys Davids and S. \V. Bushell. 2 vols.
London 1904 '5.
Relation dis voyages faits par Ics Araies et Us Pcisans daits I' hide et a
la Chine dans le IX' siccle de I'ere chretienne. Texte arabe, imiirimc
en 181 1 par les soins de feu Langles, public avec des corrections
et additions et accompagne d'une traduction franqaise et d'cclaircisse-
ments, par J. T. Reinaud. 2 tomes 16° Paris 1845.
Marco Polo. The book of Ser Marco Polo, the Venetian, concerning the
kingdoms and marvels of the East. Translated and edited, ^with notes,
by H. Yule. London 1871; — a""* ed. 1875; — 3'* ed., revised by
H. Cordier. 2 vols. London 1903.
Friar Jordanus. Mirabilia descripta, the wonders of the East. Translated
from the Latin original as published at Paris in 1839, in the "Recueil
de Voyages et de Mcmoires" etc., by Col. H. Yule. London 1863
(Hakluyt-Society, vol. 31).
India in the !§"• century, being a collection of narratives of voyages of
India in the century preceding the Portuguese discovery of the Cape
of Good Hope, from Latin, Persian, Russian, and Italian sources, now
first translated into English. Ed. with an introduction by R. H. Major.
London 1857 (Hakluyt-Society, vol. 22).
Maffei, J. r. Selectarum epistolarum ex India libri quatuor, Joanne Petro
Maffeio interpretc. 4° Venetiis 1588; — id. 4° Viennae Austriae 1751; —

Historiarum Indicarum libri XVI. 4° Florcntia 1588; — id. 4° Viennae
Austriae 1751 ; —

Le istorie dell' Indie orientali. Tradotte di latino in lingua toscana da Fran-
cesco Serdonati. 40 Venezia 1589; — new ed. 3 vols. Milano 1S06: —

Historiarum Indicarum libri XVI. Selectarum item ex India epistolarum
codem interprcte libri IV. Access. Ign. Loiolae vita postremo recogn.
4° Venetiis 1589; —

J. P. Maffeii historiarum Ind. libri XVI. Selectarum item ex India epist.
libri IV. Accessit liber rccentiorum epist. a Joanne Hayo Dalgattiensi

Travels. i 79

Scoto ex eadem societate nunc primum excusus. 2 partes Antuerpiae
1605 ; — L'histoire des Indes orientales et occidentales. Paris 1669.

LiNSCHOTEN, HuYGHEN VAN. Itincrario, voyage of te schipvaert . . . van Jan
Huygen van Linschotcn. 4° To t'Amstelrcdam 1596; —
Navigatio ac itincrarium Joh. Hug. Linscotani in oricntalem sive Lusi-
tanorum Indiam. Dcscriptiones eiusdem terrac. Imagines, habitus
gestusque Indorum ac Lusitanorum per Indiam vivcntium, tcmplorum,
idolorum, aedium, arborum, fructuum, etc. Mores gentium circa sacri-
ficia, politiam ac rem familiarem. Collecta omnia ac descripta per eundem
belgice, nunc vero latino reddita. 2 partes fol. Hagae-Comitis 1599; —
John Hoighen van Linschoten, his discours of voyages into yc Easte
and VVeste Indies. Transl. by Wm. Phillips, fol. London 1598; —
The first book of the voyage of John Huyghen van Linschoten to the
East Indies, from the old English translation of 1598; ed. by Burnell
and Tiele. 2 vols. London 1885 (Hakluyt-Society, vols. 70 — 71); —
Histoire de la navigation . . . de Jean Hughes de Linscot et de son
voyage aux Indes orientales, avec annotations de Bernard Paludanus.
2= ed. 4° Amsterdam 1619; — 3= cd. fol. Amsterdam 1638.

De Brv, J. T. AND J. I. Collectiones peregrinationum in Indiam orientalem.
Pars I— VIII. fol. Francofurti 1598— 1607.

Hakluvt, R. The principal navigations, voiages, traffiques, and discoveries
of the English nation . . . Reprinted from the rare ed. of 1599,
with additions. 5 vols. 4" London 1809 — 12; — new ed. 12 vols.
Glasgow 1903/5.

CovERTE, R. A true and almost incredible report of an Englishman, that
— being cast away in the good ship called the Assension in Cambaya,
the farthest part of the East Indies — travelled by land through many
unknowne kingdomes, and great cities ... As also a relation of their
commodities and manner of traffique, and at what seasons of the
yeere they are most in use . . . With a discovery of a great empcrour
called the Great Mogoll, a prince not till now knowne to our English
nation. London 1612; — 4° London 1631.

Roe, Sir Th. The embassy of Sir Th. Roe to the court of the Great Mogul,
161 5 — 19, as narrated in his journal and correspondence. Ed. by
W. Foster. 2 vols. London 1899 (Hakluyt-Society, 2"d series, vols. 1—2).

Wheeler, J. T. Early travels in India: i'^' series, from Purchas's pilgrimage
and Linschoten's voyages. Calcutta 1864; 2"^ scries, comprising Roe's
embassy and Fryer's travels. London 1873.

Valle, Pietro della. Viaggi di Pietro della Valle il pellegrino. Descritti
da lui medesimo in 54 lettere familiari. Divisi in tre parte, dot la
Turchia, la Persia e I'lndia. 3 vols. 4° Rom 1650 8; —
The travels of Sig. Pietro della Valle, a noble Roman, into East India
and Arabia deserta. In which the several countries, together with
the customs, manners, traffique and rites, both religious and civil, of
those oriental princes and nations, are faithfully described, in familiar
letters to his friend Signior Mario Schipano. Translated by G. Havers.
Whereunto is added a relation of Sir Thomas Roe's voyage into the
East Indies, fol. London 1665 ; — The travels of Pietro della Valle
in India, from the old English translation of 1664, by G. Havers;
ed. with a life of the author, and introduction and notes, by Edward
Grey. 2 vols. London 1892 (Hakluyt Society, vols. 84, 85); —

•So 5. Ethnography.

Les fameux voyages de Pietro della Valle . . . 3 tomes 4" Paris 1670/84;
— nouv. cd. 8 tomes 12° Paris 1745.
Dapper, O. Asia, of naukeurige bcschryving van het rijk des grooten
Mogols, en een groot gedeelte van Indicn, behelsende de landschappen
van Kandahar, Kabul, Multan, Haikan, Bukkar, Send of Din, Jesselmeer,
Attak, Peniab, Kaximir, Jangapore, Dely, Mando", Malva, Chitor, Utrad,
Zuratte of Kambaye, Chandisch, Narvar, Gwaliar, Indostan, Sonbat,
Bakar, Nagrakat, Dekan en Visiapour . . . fol. Amsterdam 1672.
Bermier, F. Voyages de Francois Bernier, cont, la description des ctats
du Grand Mogol, de I'Hindoustan, du royaume de Kachemirc, etc.

2 tomes Amsterdam 1699; — do., 1710; do., 1724;

Travels in the Mogul Empire, A. D. 1656 '68. History of the late revo-
lution of the empire of the Great Mogol . . . London 1671 ; — reprinted,
Bombay 1830; — Calcutta (i860); — Calcutta (1904); —
Travels in the Mogul Empire. Transl. from the French by Irving Brock.
2 vols. London 1826; — revised and improved by A. Constable.
Westminster 1891.
Tavernier, J. B. Les six voyages en Turquie, en Perse, et aux Indes.
2 tomes 4° Paris 1676; — 3 tomes 12" (Elzevier) 1679; — 6 tomes
I2» Paris-Rouen 1713-24; — 3 tomes La Haye 17 18; —
English translation fol. London 1678; — fol. London 1684: — Travels
in India. Translated from the original French ed. of 1676, with a
biographical sketch of the author, notes, appendices etc., by V. Ball.
2 vols. London 1889; —
Beschreibung der 6 Reisen in Tiirckey, Persian und Indien innerhalb
40 Jahren . . . Deutsch von J. H. Widerhold. fol. Genf 1681 ; —
Vierzigjahrige Reisebeschreibung . . ., sechsmalige Landerreise durch
Tiirckey, Persian, Indien . . . Deutsch durch J. Menudier. fol. Nurn-
berg 16S1 ; — 2. A. 1690.
Fryer, J. A new account of East India and Persia, in 8 letters; being
9 years' travels, 1672—81, containing observations made of the moral,
natural, and artificial state of those countries; namely, of their govern-
ment, religion, etc. fol. London 1698; — new ed., by W. Crooke.
Vol. I. London 1909 (Hakluyt-Society, Series II, vol. 19).
Grose, J. H. A voyage to the East Indies, with observations there (1750—56).
London 1757; — A voyage to the East Indies, with observations con-
tinued to 1764, including authentic accounts of the governments of
the Deckan and Bengal, etc. 2 vols. London 1766; — A voyage to
the East Indies; containing authentic accounts of the Mogul govern-
ment in general, the viceroyalties of the Decan and Bengal, with their
several subordinate dependances; of Angria, the Morattoes, and Tan-
joreans; of the Mahometan, Gentoo, and Parsee religions; of their
customs and antiquities, with general reflections on the trade of
India, etc. 2 vols. London 1772.
SoNNERAT. Voyage aux Indes orientales et a la Chine, 1774—81. 2 tomes
4" Paris 1782; — (another ed., 3 tomes 8° Paris 1782, is without
the plates); — nouvellc cd., augmentee, par Sonnini. 4 tomes 12"
et atlas 4" Paris 1806; — do., Paris 1836; —
Reise nach Ostindien und China.. . 2 Bde. 4° Zurich 17S3; —
Voyage to the East Indies and China. Calcutta 1788—9.

Travels. i8i

FoRSTER, G. Journey from Bengal to England through Northern India,
Kashmire, Afghanistan, etc., 1783 — 84. Calcutta 1790; — new ed.

2 vols. 4" London 179S; — new ed. 2 vols. 8° London 1808; —
Reisc aus Bengalen nach England durch Hindustan, Kaschmir, Afga-

nistan, Persien und Rufiland. Deutsch von Meiners. 2 Bde. Zurich
1796— 1800; —
Voyage du Bengale a Petersbourg. Traduit par Langlts. 3 tomes
Paris 1802.

Hodges, W. Travels in India, 1780 — 83. 4° London 1793; — 2"'' ed.,
corrected. 4" London 1794.

Paolino DA S. Bartolomeo. Viaggio alle Indife Orientali. 4° Roma 1796; —
Des Fra. Paolino da S. Bartolomeo Reise nach Ostindien. Herausgeg. von

J. R. Forster. Berlin 1798; —
A voyage to the East Indies ; with notes by J. R. Forster. Transl. from

the German by W. Johnston. London 1800 ; —
Voyage aux Indes orientates, traduit par M(arches) ; avcc des obser-
vations par Anquetil du Perron, J. R. Forster, et Silvestre de Sacy.

3 tomes Paris 1808.

Heber, Bishop R. Narrative of a journey through the upper provinces of
India, 1824 — 25, with notes upon Ceylon, journey to Madras and
southern provinces in 1S26, and letters written in India. 2 vols. 40
London 1828; — 3'' ed. 3 vols. 8° London 1828; — 4''' ed. do.,
London 1829; — new ed. 2 vols. London 1844; —
Voyage a Calcutta, a Bombay et dans les provinces superieures de I'Inde.

2 tomes Paris 1830; —
Bericht iiber eine Reise durch die oberen Provinzen von Vorderindien.
2 Bde. Weimar 1831/2.

JACQUEMONT, V. Voyage dans I'Inde pendant les annees 1828 — 32. 6 tomes
(2 tomes planches) 4° Paris 1841 — 46.

Orlich, L. v. Reise in Ostindien in Briefen an A. v. Humboldt und C. Ritter.
40 Leipzig 1845; —
Travels in India, including Sinde and the Punjaub. Transl. from the
German by H. E. Lloyd. 2 vols. London 1845.

Waldejiar, Prinz von Prelssen. Zur Erinnerung an die Reise nach Indien,
1844 — 46. Mit Vorwort von A. von Humboldt. 2 Bde. fol. Berlin 1853.

Graul, K. Reise nach Ostindien iiber Palastina und Egypten, 1849 — 53.
5 Bde. Leipzig 1854 — 56.

ScHLAGiNTWEiT, E. Reisen in Indien und Hochasien. Eine Darstellung
der Landschaft, der Cultur und Sitten der Bewohner usw. 4 Bde.
Jena 1869 — So.

RoussELET, L. L'Inde des Rajahs. Voyage dans I'Inde centrale et dans
les presidences de Bombay et du Bengale. fol. Paris 1875; — 2« dd.
4" Paris 1877; —
India and its native princes: travels in Central India and in the presi-
dencies of Bombay and Bengal. Revised and edited by Lieut.-Col.
Buckle. 4° London 1875; — 2"^ ed. 1876; — 3'' ed. 1878.

D.^HLM.\NN, J. Indische Fahrten. 2 Bde. Freiburg i. Br. 1908.

Oaten, E. F. European travellers in India during the 15''', iG'!", and 17'''
centuries. London 1909.

'82 5- Ethnography.

The people of India.

Latham, R. G. Ethnology of India. London 1859.

Watson, J. F., and J. W. Kaye. The people of India. A series of 468
photographic illustrations, with descriptive letterpress, of the races
and tribes of Hindostan. 8 vols. 4° London 1868 — 75.

RowNEV, H. B. The wild tribes of India. London 1882.

Shoshee Chunder Dutt. The wild tribes of India. London 1884.

Mac Ritchie, D. Account of the Gypsies of India. London 1886.

Oppert, G. On the original inhabitants of Bharatavarsha or India. Madras-
London i888,'9; — 2"d ed. Madras-Westminster 1893.

The principal nations of India. Madras 1S90; — 2"<* ed. Madras 1892.

Mesroob Seth, J. History of the Armenians in India, from the earliest
times to the present day. London 1S97.

Bonarjee, p. D. a handbook of the fighting races of India. Calcutta 1899.

RisLEY, H. H. The people of India. With 8 appendices. Calcutta 1908.

Menpes, M. The people of India. London 1910.

Castes.

Ramaswajiy, C. V. A digest of the different castes of India, with accounts
of them. Madras 1837.

Rodriguez, E. A. The Hindoo castes ; the history of Braminical castes,
containing a minute description of the origin, ceremonies, idolatry,
manners, customs, etc., of the forty-two sects of Bramins of the
British-Indian empire. 4" Madras 1846.

Wilson, H. H. Sketch of the religious sects of the Hindus. Calcutta 1846.

Sherring, M. a. Hindu tribes and castes. 3 vols. 4° Calcutta, Bombay,
London 1872—81.

Kitts, E. J. A compendium of the castes and tribes found in India. Com-
piled from the census reports for the various provinces (excl. Burmah)
and native states of the empire, fol. Bombay 1885.

Oman, J. C. The mystics, ascetics, and saints of India. A study of Sadhuism,
with an account of the Yogis, Sanyasis, Bairagis, and other strange
Hindu sectarians. London 1903; — 2'"^ ed. 1905.

MoTi Lal Seth. A brief ethnological survey of the Khattris. Agra 1905.

Theory of caste.

Bower, Rev. H. An essay on Hindu caste. Calcutta 1851.

Irving, B. A. Theory and practice of caste in India. 12° London 1853.

Patter.son, a. J. Caste considered under its moral, social, and religious

aspects. London 1861.
Esquer. Castes dans I'lnde. Pondichery 1870.

Wilson, Rev. J. Indian caste. 2 vols. Edinburgh, London, Bombay 1877.
Hopkins, E. W. The mutual relations of the four castes, according to the

Manavadharmacjastram. (Diss.) Leipzig 1 88 1.
Jogendra Nath Bhattacharya. Hindu castes and sects; an exposition

of the origin of the Hindu caste system and the bearing of the sects

towards each other and towards religious systems. Calcutta 1896.
Senart, E. Les castes dans I'lnde; les faits et le systf^me. Paris 1S96.

Theory of Caste. — The Thugs. — Manners and customs, etc. 183

Chentsal Rao, P. The principles of Pravara and Gotra. a"*" cd., revised.

Mysore 1900.
Narayan Chandra S.aha. The Vaishya Khondosaha and the Shaundika,

and the commentary on Hindu castes. Calcutta 1907.
BouGLE, C. Essais sur le regime des castes. Paris 1908.

The Thugs.

Sleeman, \V. H. Ramaseeana, or, a vocabulary of the peculiar language
used by the Thugs, with an introduction and appendix, descriptive
of the system pursued by that fraternity and of the measures which
have been adopted by the supreme government of India for its sup-
pression. Calcutta 1836.

(Thornton, E.] Illustrations of the history and i)ractices of the Thugs,
and notices of some of the proceedings of the government of India,
for the suppression of the crime of Thuggee. London 1837.

Sleem.an, W. H. The Thugs, or Phansigars of India; a history of the
assassins and their system. 2 vols. Philadelphia 1839.

Taylor, M. Confessions of a Thug. 3 vols. London 1840; — new ed.
London 1890.

Hutton, J. A popular account of the Thugs and Dacoits, the hereditary
garotters and gang-robbers of India. London 1857.

Hervey, Ch. Operations in the Thuggee and Dacoity department, 1859 — 60.
Calcutta 1 861 (Indian Govt, records, No. 34).

Manners and customs, rites and ceremonies.

Abul Fazl: Ayeen.-\.kbery; or, the institutes of Akber, transl. by F. Gladwin.
2 vols. 4° London 1 800; —
The Ain-i-Akbari, from the Persian by H. Blochmann. 2 vols. Calcutta

1873—91.
Lord, H. A display of two forraigne sects in the East Indies, viz., the

sect of the Banians, the ancient natives of India, and the sect of the

Parsees, the ancient inhabitants of Persia; together with the religion

and manners of each sect, etc. London 1630 ; — another ed. fol.

London 1747; —
Histoire de la religion des Banians, avec un traite de la religion des

Parsis. Paris 1667.
Palladils. De gentibus Indiae et Bragmanibus; graece et latine. S. Am-

brosius, de moribus Brachmanorum; latine. Anonymus, de Bragmanibus ;

latine. Nunc primum edidit Ed. Bissaeus. 4° Londini 1665.
[Craufurd, Q.] Sketches chiefly relating to the history, religion, learning,

and manners of the Hindoos. London 1790: — 2"'^ ed., enlarged.

2 vols. London 1792.
SoLVYNS, B. A collection of two hundred and fifty coloured etchings, des-
criptive of the manners, customs, and dresses of the Hindoos. 2 vols.

fol. Calcutta 1799; —
The costumes of Hindustan, elucidated by sixty coloured engravings;

with descriptions in English and French, taken in the years 1798

and 1799. 4° London 1804; —
Les Hindous, ou description de leurs moeurs, coutumes, ceremonies, etc.,

dessines d'aprcs nature dans le Bcngale, et representes en 292 planches;

avec le te.xte en frangais et en anglais. 4 tomes fol. Paris 1808 — 12.

1 84 5. Ethnography.

Moor, E. The Hindu pantheon. 4" London 1810; — new ed., condensed
and annotated by W. O. Simpson. Madras 1864.

Moor, E. Hindu infanticide; an account of the measures adopted for
suppressing the practice of the systematic murder by their parents
of female infants ; with incidental remarks on other customs peculiar
to the natives of India. 4° London 181 1.

Ward, W. Account of the writings, religion, and manners of the Hindoos.
4 vols. 4° Serampore 181 1 ; — (2"'' ed.) A view of the history, lite-
rature, and mythology of the Hindoos. 2 vols. 4° Serampore 1815;

— S"" ed. 4 vols. 8° London 1817 — 20; — new i^"") ed. 3 vols.
London 1822; — $'^ ed., abridged and greatly improved, with biogr.-
sketch of the author and an ample index. Madras 1863.

D'Oyley, Sir Charles. The costume and customs of modern India ; from
a collection of drawings. With a preface and copious descriptions
by Capt. Th. Williamson. 4° London (1813).

Dubois, J. A. Description of the character, manners, and customs of the
people of India, and of their institutions, religious and civil. Transl.
from the French Ms. 4° London 1817; — another ed., with notes
and additions by Rev. G. U. Pope. Madras 1862; — reprinted from
the orig. ed. of 1817. Madras 1879; — new ed. Calcutta 1905; —
Maurs, institutions et ceremonies des peuples de I'lnde. 2 tomes

Paris 1825; —
Hindu manners, customs, and ceremonies. Transl. from the author's
later French l\Is., and ed. with notes, corrections, and biography by
H. K, Beauchamp. 2 vols. Oxford 1897; — s"* ed. Oxford-London 1906.

Malcolm, Sir John. Essays relative to the habits, character, and moral
improvement of the Hindoos. London 1823.

Coleman, Ch. The mj-thology of the Hindus, with notices of various moun-
tain and island tribes, and an appendix comprising the minor Avatars,
and the mythological and religious terms, etc. 4° London 1832.

Meer Hassan Ali, Mrs. Observations of the Mussulmanns of India, des-
criptive of their manners, customs, habits, and religious opinions,
made during a twelve years' residence in their immediate society.
2 vols. London 1832.

Qanoon-e-Islam^ or, the customs of the Moosulmans of India; comprising
a full and exact account of their various rites and ceremonies. By
Jaffur Shurreef, a native of the Deccan. Composed under the direction
of, and transl. by G. A. Herklots. London 1S32; — 2°'* ed. iSIadras 1S63;

— new ed. Madras 1895.

Bjornstjerna. The theogony of the Hindoos. London 1S44.

Belnos, Mrs. S. C. The Sundhya, or, the daily prayers of the Brahmins.
Illustrated in a series of original drawings from nature, together with
a descriptive text, etc. fol. Allahabad 1851.

Wilson, Rev. J. History of the suppression of infanticide in Western India
under the government of Bombay; including notices of the provinces
and tribes in which the practice has prevailed. Bombay 1855.

Cave-Browne, Rev. J. Indian infanticide; its origin, progress, and sup-
pression. London 1857.

Wilson, H. H. Essays and lectures on the religion of the Hindus. Ed. by
Dr. R. Rost. 2 vols. London 1862.

Manners and customs, rites and ceremonies. 185

Wilson, H. H. Hindu religions, or, an account of the various religious
sects of India. Calcutta 1899.

West, R., and G. BChler. Digest of Hindu law, from the replies of the
Shastris in the several courts of the Bombay presidency. 2 vols.
Bombay 1867/9.

Garrett, J. A classical dictionary of India, illustrative of the mythology,
philosophy, literature, antiquities, arts, manners, customs, etc., of the
Hindus. Madras 1871/3.

Pratap Chandra Ghosha. Durga Puja, with notes and illustrations. Cal-
cutta 1 87 1.

Pratap Chandra Ghosha. Origin of the Durga Puja. Calcutta 1874.

JIONiER-Wii-LUMS, M. Indian wisdom, or, examples of the religious, philo-
sophical, and ethical doctrines of the Hindus. London 1 87 5; — 3'' ed.
1876; — 4''' ed. 1893.

KiTTEL, F. Uber den Ursprung des Linga-Cultus in Indien. Mangalore 1876.

Banerjee, G. The Hindu law of marriage and Stridhana (being the Tagore
law lectures for 1878). 2"'' cd., revised. Calcutta 1896.

Monler-Williams, M. Modern India and the Indians, being a series of
impressions, notes, and essays. London 1878; — s*" ed. 1879; —
4'i> ed. 1887; — 5'h ed.

Sanderson, G. P. Thirteen years among the wild beasts of India; their
haunts and habits from personal observation; with an account of
the modes of capturing and taming elephants. 8° London 1878; —
3d ed. 1882; — 6"> ed. 4° 1896.

Barth, a. Les religions de I'lnde. Paris 1879; —

The religions of India. Transl. by Rev. J. Wood. London 1882.

DowsoN, J. A classical dictionary of Hindu mythology and religion,
geography, history, and literature. [London 1879.

Momer-Williams, M. Religious thought and life in India. An account of
the religions of the Indian peoples. Part I: Vedism, Brahmanism,
and Hinduism. London 1883; — 3<i ed. : Brahmanism and Hinduism,
or, religious thought and life in India. London 1887; — 4''' ed.,
enlarged. London 1S91.

Shoshee Chunder Dltt. Realities of Indian life; or, stories collected
from the criminal reports of India, to illustrate the life, manners, and
customs of its inhabitants. London 1885.

JIalabarj, B. M. Infant-marriage and enforced widowhood in India, fol.
Bombay 18S7.

Om.\n, J. C. Indian life, religious and social. London (Philadelphia) 1889.

KiST, L. Indisches, und zwar: Religion, Tempel und Feste der Hindu.
Innsbruck 1S90.

Lyall, Sir A. C. Natural religion in India. Cambridge 1891.

Ramakrisna, Y. Life in an Indian village. With an introduction by Sir
M. E. Grant Duff. London 1891.

Scherman, L. Materialien zur Geschichte der indischen Visionslitteratur.
Leipzig 1892.

BiLLiNGTON, Miss M. F. Woman in India. London 1895.

P.\DFiELD, J. E. The Hindu at home; account of domestic rites and cere-
monies. London 1895; — 2"<i ed. Vepery -Madras 1908.

Hopkins, E. W. The religions of India. London 1896.

1 86 5. Ethnography.

Mangaldas Nathlbov. Lectures on Hindu castes, ceremonies, customs,
and inheritance. Bombay 1896.

Schmidt, R. Beitragc zur indischen Erotik. Leipzig 1902; — 2. Aufl.
Berlin 1910.

Sellon, E. Annotations on the sacred writings of the Hindus, being an
epitome of some of the most remarkable and leading tenets in the
faith of that people, illustrating their priapic rites and phallic prin-
ciples. New ed. London 1902.

Natesa Sastri, S. 1\I. Hindu feasts, fasts, and ceremonies. With an in-
troduction by H. K. Beauchamp. Madras 1903.

Murdoch, J. Hindu and Muhammadan festivities. Compiled from Wilson,
Wilkins, Crooke, Sell, Hughes, and other writers. Madras 1904.

R.AM Sahay Mukherji. Indian folklore. Calcutta 1904; — 2"<' ed. Cal-
cutta 1906.

Schmidt, R. Liebe und Ehe im alten und modernen Indien. Berlin 1904.

Oman, J. C. The Brahmans, Theists, and Muslims of India. Studies of
goddess-worship in Bengal; caste, Brahmaism, and social reform;
with descriptive sketches of curious festivals, ceremonies, and faquirs.
London 1907; — 2"'^ ed. 1909.

Schmidt, R. Fakire und Fakirtum im alten und modernen Indien. Yoga-
Lehre und Yoga-Praxis. Nach indischen Originalquellen. Berlin 1907.

Natesa Sastri, S. M. Indian folk-tales. With an introduction by A. G.
Cardew. 2"^ ed., enlarged. Madras 190S.

Oman, J. C. Cults, customs, and superstitions of India. Beiftg a revised and
enlarged edition of "Indian life, religious and social". London 1908.

R.agunatha Rao, R. The Aryan marriage, with special reference to the
age-question: a critical and historical study. Madras 1908.

Coomarasvam\-, a. K. The Indian craftsman. London 1909.

Totemism in India: see Vol. II, pp. 218—335 of: J. G. Frazer, Totemism
and exogamy, a treatise on certain early forms of superstition and
society. 4 vols. London 1910.

Art-manufactures and products.

Watson, J. F. The textile manufactures and the costumes of the people
of India. 4° London 1866; — another ed. fol. London 1867.

Watson, J. F. Index to the native and scientific names of Indian and other
Eastern economic plants and products. London 1868.

Watson, J. F. Collection of .specimens and illustrations of the textile manu-
factures of India. 2'"' scries. 17 vols. fol. and 4" London 1873—77.

Birdwood, G. C. M. The industrial arts of India. 2 parts. London 1880;
— new ed. 1884.

Indian art-manufactures. 51 photographic illustrations, taken by order of
the government of India of some selected objects shown at the 3''
exhibition of native fine and industrial art at Simla iSSi. With report
by H. H. Cole. 4° [Simla] 18S3.

Mukharji, T. N. a handbook of Indian products, art-manufactures, and
raw materials. Published by J. Patterson. Calcutta 1883.

Mukharji, T. N. Art-manufactures of India. Specially compiled for the
Glasgow international exposition 1888. Calcutta i888.

Arts-manufactures and products. — Journals. 187

Watt, Sir George. A dictionary of the economic products of India.

Vol. I — VI and index. 10 vols. London-Calcutta 1889—93.
Preservation of the national monuments of India. (lOO plates of the industrial

and ornamental arts.) 4° London 1896.
Watt, .Sir George. The commercial products of India ; being an abridgment

of "The dictionary of the economic products of India". London 1908.

Journals.

Asiatick Researches, or transactions of the Society, instituted in Bengal
for inquiring into the history and antiquities, arts, sciences, and lite-
rature of Asia. Vol. I — XX. 4" Calcutta 1788— I S36; — reprinted, 4"
London 1799 sqq.; — reprinted, 8° Vol. I — XII. London 1798 — 1818.
(Index to vols. I— XVIII. 40 Calcutta 1835. Index to vols. XIX— XX
of Asiatic Researches, and vols. I — XXIII of the Journal of the Asiatic
Society of Bengal. Calcutta 1856.)

Journal of the Asiatic Society of Bengal. Vols. I— LXXIII. Calcutta
1832 — 1904. (Part II: Natural history, to vols. 34 — 73; Part III: An-
thropology, to vols. 62 — 73.)

Proceedings of the Asiatic Society of Bengal. Calcutta 1865— 1904.

Journal and proceedings of the Asiatic Society of Bengal. Vol. I etc.
Calcutta 1905 sqq.

Memoirs of the Asiatic Society of Bengal. Vol. I etc. 4° Calcutta 1905 sqq.

Transactions of the Royal Asiatic Society of Great Britain and Ireland.
3 vols. 4° London 1827 — 35.

Journal of the Royal Asiatic Society of Great Britain and Ireland.
London 1834 sqq.

Journal of the Bombay Branch of the Royal Asiatic Society. Bombay 1 841 sqq.

Journal of the Ceylon Branch of the Royal Asiatic Society. Colombo 1845 sqq'.

Asiatic Journal, and monthly register for British India and its depen-
dencies. London i8i6sqq.; — (^Selections from vols. I — XXVIII. 2 vols.
Madras 1875.)

Transactions of the Literary Society of Bombay. 3 vols. 4° London
1819—23; — reprinted, 3 vols. Bombay 18767.

The Madras Journal of literature and science ; published under the auspices
of the Madras Literary Society and auxiliary of the Royal Asiatic Society.
Ed. by J. C. Morris, R. Cole, C. D. Brown, etc. Madras 1834 sqq.

The Calcutta Review. Calcutta 1 844 sqq.

The Indian Antiquary. 4° Bombay 1872 sqq.

The Bengal Magazine. Ed. by Lai Behari Day. Calcutta 1873 sqq.

The Journal of Indian Arts (from vol. VII : ... Art and Industry)- fol.
London 1883 sqq.

The Madras Christian College Magazine. Madras 1883 sqq.

The Orientalist. A journal of oriental literature, arts and sciences, folk-
lore, etc. Ed. by W. Goonetilleke. Kandy 1885 sqq.

Journal of the Anthropological Society of Bombay. Bombay 1895 sqq.

The New Asiatic Rez'iew. Ed. by C. M. Barrow and J. Collyer Adam. 4°
Madras 1907 sqq.

Journal Asiatique. Public par la Societe Asiatique. Paris 1822 sqq.

l88 5. Ethnography.

Zcitschrift fur die Kunde des Morgenlandcs. Herausgeg. von Ewald, v. d.

Gabelentz, Kosegarten, Lassen, etc. 7 Bde. Gottingen undBonn

1837— SO.
Zcitschrift der Dciitschen Morgenlandischen Gesellsckaft. Leipzig 1850 sqq.
Wiener Zcitschrift fiir die Kunde des Morgenlandcs. Wien 1887 sqq.

Northern India, general.

Ha.milton, Ch. An historical relation of the origin, progress, and final
dissolution of the government of the Rohilla Afgans in the northern
provinces of Hindostan, etc. London 1787; — 2"<' ed. London 1788.
Tod, J. Travels in Western India, embracing a visit to the sacred mounts
of the Jains, and the most celebrated shrines of Hindu faith between
Rajpootana and the Indus; with an account of the ancient city of
Nehrwalla. 4° London 1839.
Sleeman, W. H. Rambles and recollections of an Indian official (including
the authors journey from the banks of the Nerbudda to the Himalayas,
1835—36). 2 vols. London 1844; — republished by A. C. Majumdar!
2 vols. Lahore 1888; — new ed., by V. A. Smith, with memoir, notes,'
index, and a bibliography. 2 vols. Westminster 1S93.
Thornton, E. A gazetteer of the countries adjacent to India on the north-
west, including Sinde, Afghanistan, Beloochistan, the Punjab, and the
neighbouring states. 2 vols. London 1844.
Thornton, E. Gazetteer of the territories under the government of the
East India Company, and of the native states on the continent of
India. 4 vols. London 1854; — 2"'i ed. London 1857 8; — revised
by Sir Roper Lethbridge and A. N. Wollaston. London 1886.
Muir, J. Notes on a trip to Kedarnath and other parts of the sno\\7
range of the Himalayas. With some account of a journey from Agra
to Bombay in 1854. Edinburgh 1855.
UjFALVY, K. E. voN. Aus dem westlichen Himalaya. Erlebnisse und

Forschungen. Leipzig 18S4.

Fallon, S. W. Dictionary of Hindustani proverbs, including many Marwari,

Panjabi, Maggah, Bhojpuri and Tirhuti proverbs. Edited and revised

by R. C. Temple and Lala Faqir Chand. Benares 1884; — 2"'^ ed. 1886.

Temple, R. Journals kept in Hyderabad, Kashmir, Sikkim and Nepal.

Ed., with introduction, by R. C. Temple. 2 vols. London 1887.
WiLKLNs, W. J. Modern Hinduism, being an account of the religion and life
of the Hindus in northern India. London-Calcutta 1S87; — 2"<'ed. 1900.
North- Iitdiar Notes and Queries; a monthly periodical, devoted to the
systematic collection of authentic notes and scraps of information
regarding the country and the people. Ed. by W. Crooke. Vols. I— V.
4° Allahabad 1891 '96.
Rislev, H. H. The tribes and castes of Bengal, North- Western Provinces,
and Punjab. I. Anthropometric data ; II. Ethnographic glossary. 4 vols.
Calcutta 1891/2.

Crooke, W. Introduction to the popular religion and folklore of northern
India. Allahabad 1894; — 2"^ ed. : Popular religion and folklore of
northern India. 2 vols. Westminster 1896.

Waddell, L. a. Among the Himalayas. Westminster 1899.

Northern India, general. — Afghanistan. 189

Sherring, C. a. Western Tibet and the British borderland. The sacred
country of Hindus and Buddhists. With an account of the govern-
ment, religion, and customs of its peoples, etc. London 1906.

Crooke, W. Natives of northern India. London 1907.

Afghanistan.

Klaproth, J. VON. Obcr die Sprache und den Ursprung der Aghuan oder

Afghanen. 4° St. Petersburg 1810.
Elphinstone, Mountstuart. An account of the Kingdom of Caubul, and

its dependencies in Persia, Tartary, and India; comprising a view of

the Afghaun nation, and a history of the Dooraunce monarchy. 4°

London 1815; — new and revised ed. 2 vols. 8" London 1839;

(3'' ed.) 2 vols. London 1842.
(Tableau du royaume de Caboul etc. ; traduit par M. Breton. 3 tomes

12° Paris 1817.)
Leech, M.aj. R- An account of the early Abdalees. Calcutta (1840?).
Wilson, M. H. Ariana antiqua, a descrijitive account of the antiquities

and coins of Afghanistan, with a memoir of the buildings called Topes

by C. Masson. 4° London 1841.
BuRNES, A. Cabool, narrative of a journey to, and residence there 1836 — 38.

London 1S42; — 2°^ ed. London 1843.
(Kabul. Schilderung einer Reise nach dieser Stadt und des Aufenthaltes

daselbst 1836 — 38. Aus dem Englischen von Th. Oclkers. Leipzig 1843.)
Hart, Capt. L. W., and J. Atkinson. Character and costumes of Afghau-

nistan. fol. London 1843.
Willis, Capt. Character and costumes of Afghanistan, fol. London 1843.
Masson, Ch. Legends of the Afghan countries, in verse, original and

translated. London 1848.
Rattray, J. The costumes of the various tribes, portraits of ladies of rank,

celebrated princes and chiefs, views and interiors of the cities and

temples of Afghaunistaun, from original drawings, fol. London 1848.
Relations of the Government with tribes on the N. IV. Frontier, 1849—55.

Calcutta 1856 (Ind. Govt. Records, No. 12).
LuMSDEN, H. B. The mission to Kandahar, with appendices and suppl.

report of the expedition into Upper Meeranzye and Koorrum in 1S56.

Calcutta i860.
Bellew, H. W. Journal of a political mission to Afghanistan, in 1857,

under Major Lumsden; with an account of the country and people.

London 1862.
Bellew, H. W. A general report on the Yusufzais, in 6 chapters. Lahore 1 864.
Mac Gregor, C. M. A contribution towards the better knowledge of the

topography, ethnology, resources, and history of Afghanistan. Calcutta

1871 (Govt, reports, Central Asia, part II).
MuH.\iniAD Hay.\t Kh.Xn. Afghanistan and its inhabitants. Translated from

the "Hayat-i-Afghan", by H. Priestley. Lahore 1874.
Thorburn, S. S. Bannu, or, our Afghan frontier. London 1876.
Bellew, H. W. Afghanistan and the Afghans, being a brief review of the

history of the country, and account of its people. London 1879.

I go 5. Ethnography.

Bellew, H. W. The races of Afghanistan, being a brief account of the
principal nations inhabiting that country. Calcutta 1880.

CoLQUHOUN, J. A. S. With the Kurram field force, 1878 — 79. L<jndon 1881.

RosKOSCHNY, H. Afghanistan und seine Nachb^rlander. 2 Bande 4"
Leipzig 1885.

Raverty, H. G. Notes on Afghanistan and part of Baluchistan, geogra-
phical, ethnographical, and historical, extracts from the writings of
little known Afghan and Tajzik historians etc., and from personal
observations, fol. London 1888.

Yate, a. C. Travels with the Afghan Boundary Commission. London 1887.

Yate, Maj. C. E. Northern Afghanistan ; letters from the Afghan Boundary
Commission. London 1888.

Bellew, H. W. An inquiry into the ethnography of Afghanistan, prepared
for and presented to the 9''' International Congress of Orientalists.
London 1891.

Murray, J. W. Dictionary of the Pathan tribes on the north-west frontier
of India. Calcutta 1899.

Sultan Muham.mad Khan. The constitution and laws of Afghanistan.
London 1900.

Warburton, Sir R. Eighteen years in the Khyber, 1879 — 98, with remi-
niscences of service among the tribes of the N.-W. Frontier of India.
I-ondon 1900. <.

Hamilton, A. Afghanistan. London 1906.

Pennell, D' T. L. Among the wild tribes of the Afghan frontier ; a record of
sixteen years' close intercourse with the natives of the Indian marches;
with introduction by Field Marshal Earl Roberts. London 1909.

Baluchistan.

Pottinger, H. Travels in Beloochistan and Sinde, with geographical and
historical account of those countries. 4° London 18 16.

Masson, Ch. Narrative of various journeys in Balochistan, Afghanistan,
and the Panjab; including a residence in those countries from 1826
to 1838. 3 vols. London 1842.

Ma-sson, Ch. Narrative of a journey to Kalat, including an account of the
insurrection at that place in 1840; and a memoir on Eastern Balo-
chistan. London 1843.

Bellew, H. W. From the Indus to the Tigris : a narrative of a journey
through the countries of Balochistan, Afghanistan, Khorassan, and
Iran, in 1872; together with a synoptical grammar and vocabulary of
the Brahoe language. London 1874.

Schlagintweit, E. Kelat: the Brahui Kingdom on the southern border
of Iran. Translated from the German of the "Ausland" by F. Henvey.
Simla 1876.

Hughes, A. W. The country of Baluchistan, its geography, topography,
ethnology and history. London 1877.

Macgregor, Col. C. M. Wanderings in Balochistan. London 18S2.

Oliver, E. E. Across the border, or, Pathan and Biloch. London 1890.

Dames, M. L. The Baloch race: a historical and ethnological sketch.
London 1904 (As. Soc. Monographs, IV).

Dames, M. L. Popular poetry of the Baloches. 2 vols. London 1907.

North -West Frontier Province. — Kashmir. 191

North -\Vest Frontier Province.

Leitner, G. W. Results of a tour in Dardistan, Kashmir, Little Tibet, Ladak,
Zanskar etc. Vol. I, part i — 3: The languages and races of Dardistan.
Comparative vocabulary and grammar of the Dardu languages. Voca-
bulary and dialogues in the Shina dialects. Legends, riddles, proverbs,
fables, customs, songs, etc., of the Shina race and the history of the
encroachments of Kashmir on Dardistan. 4° Lahore 1866 — 73.

Maclean, C. B. A compendium of Kafir laws and customs. Including
genealogical tables of Kafir chiefs and various tribal census returns.
1866; — reprinted. Grahamstown 1906.

BiDDULPH, J. Tribes of the Hindoo Koosh. Calcutta 1880.

Leitner, G. W. A sketch of the Changars, and of their dialect, fol.
Lahore 1880.

Leitner, G. W. The Hunza and Nagyr handbook. Part I : The Kajuhna
and Shina languages (vocabulary, dialogues, songs, legends, fables,
riddles, etc.). fol. Calcutta 1889.

Leitner, G. W. The Hunza and Nagyr handbook. Being an introduction
to a knowledge of the language, race, and countries of Hunza, Nagyr,
and a part of Yasin. With a supplement on Dardistan in 1866, 1886
and 1893 ... 2 vols. 4° and 8° (vol. I: 2'"' ed.). Woking 1893.

O'Brien, D. J. T. Grammar and vocabulary of the Khowar dialect (Chitrali),
with introductory sketch of country and people. Lahore 1895; —
2""' ed. Calcutta 1898.

Robertson, G. S. The Kafirs of the Hindu-Kush. London 1S96; — new ed.
London 1900.

UjFALVY, K. E. VON. Les Aryens au nord et au sud de I'Hindu-Kouch.
Paris 1S96.

Olufsen, O. Through the unknown Pamirs. The second Danish Pamir
expedition, 189S — 99. London 1904.

Ahrt/i- IVest Frontier Province district-gazetteer. (Vol. I B.) Peshawar 1907.

Kashmir.

HCgel, Baron C. von. Kaschmir und das Reich der Siek. 4 Bde. Stuttgart

1840—48. — Travels in Kashmir and the Punjab, with account of the

Sikhs; from the German with notes by Major T. B. Jervis. London 1845.
ViGNE, G. T. Travels in Kashmir, Ladak, Iskardo, the countries adjoining

the mountain-course of the Indus, and the Himalaya, north of the

Panjab. 2 vols. London 1842.
Cl-nninghaM, a. Ladak, physical, statistical, and historical; with notices

of the surrounding countries. London 1854.
B.\TES, E. C. Gazetteer of Kashmir and the adjacent districts. Calcutta

1873 (Govt, reports. Central Asia, Part VII, Section I).
Drew, F. The Jummoo and Kashmir territories. London 1875.
Wakefield, W. The happy valley, sketches of Kashmir and the Kashmiris.

London 1879.
Knowles, J. H. A dictionary of Kashmiri proverbs and sayings. Explained

and illustrated from the rich and interesting folklore of the valley.

Bombay 1885.

192 5- Ethnography.

Knowles, J. H. Folk-tales of Kashmir. London 1888; — a""" ed. 1893.
Knight, E. F. Where three empires meet: a narrative of a recent travel

in Kashmir, western Tibet, Gilgit, and the adjojning countries. London

1893; — new ed. 1897; — "ew ed. 1905.
Lawrence, \V. R. The valley of Kashmir. 4" London 1895.
PiRiE, P. Kashmir, the land of streams and solitudes. 4° London-New

York 1909.

Nepal.

KiRKiwTRicK, Col. W. An account of the Kingdom of Nepaul, being the
substance of observations made during a mission to that country in
the year 1793. 4° London 181 1.

Hamh-ton, F. An account of the kingdom of Nepal, and of the territories
annexed to this dominion by the house of Gorkha. 4" Edinburgh 1819.

Cavenagh, O. Rough notes on the state of Nepal, its government, army,
and resources. 12° Calcutta 185 1.

Smith, Th. Narrative of a five years' residence at Nepaul. 2 vols.
London 1852.

Wright, D. Sketch of the portion of the country of Nepal open to
Europeans. 4° Calcutta 1872.

H0DG.SON, B. H. Essays on the languages, literature, and religion of Nepal
and Tibet. London 1874.

Wright, D. History of Nepal, translated from the Parbatiya by Munshi
Shew Shunker Singh and Pandit Shri Gunanand ; with an introductory
sketch of the country and people of Nepal by the editor. Cam-
bridge 1877.

Oldfield, H. a. Sketches from Nipal, historical and descriptive, with
anecdotes of the court life and wild sports of the country in the
time of Maharaja Jang Bahadur. 2 vols. London i88o.

Vansittart, Capt. E. Notes on Nepal. With an introduction by H. H. Risley.
Calcutta 1896.

Levi, S. Le Nepal. Etude historique d'un royaume hindou. 3 tomes
Paris 1905/8. (Annales du Musee Guimet, Bibliothcque d'Etudes,
t. XVII— XIX.)

Sikkim and Bhutan.

Turner, S. Account of an embassy to the court of the Teshoo Lama in

Tibet; narrative of a journey through Bootan, 17S3 — 4. 4° London

1800; — 2"^ ed. 4° London 1806.
Pemberton, R. B. Report on Bootan. Calcutta 1839.
Eden, A. Political missions to Bootan, comprising the reports of the

Hon. Ashley Eden, 1864; Capt. R. B. Pemberton, 1837— 8; with Dr. W.

Griffith's journal, and the account by Baboo Kishen Kant Bose.

Calcutta 1865.
Rennie, D. F. Bhotan and the story of the Dooar war, including sketches

of a three month's residence in the Himalayas, and narrative of a

visit to Bhotan in may, 1865. London 1866.

SiKKiM AND Bhutan. — Assam. 193

Edgar, J. W. Report on a visit to Siiihim and the Thibetan frontier, 1873.

Calcutta 1874.
Macaulay, C. Report of a mission to Sikkim and the Tilictan frontier.

4° Calcutta 1885.
The Gazetteer of Sikkim; with an introduction by H. H. Risley. Edited in

the Bengal Govt. Secretariat. 4" Calcutta 1894.
Louis, J. A. H. The gates of Thibet. A bird's eye view of independent

Sikkhim, British Bhoolan, and the Dooars, as a Doorga Poojah trip.

Calcutta 1894.
DoNALD.soN, F. Lepcha land, or, six weeks in the Sikhim Himalayas.

London 1900.
White, J. C. Sikhim and Bhutan; 21 years on the North-East frontier,

1887— 1908. London 1909.
Assam.

(cp. H. CoRDiER. Bibliographic, in T'oung I'ao, Scrie 11, Vol. VII, 163 etc.
and Vol. IX, 137 etc.)

Robinson, W. A descriptive account of Asam, with a sketch of the local
geography, and a concise history of the tea-plant; to which is added
a short account of the neighbouring tribes, exhibiting their history,
manners, and customs. Calcutta 1841.

Owen, J. Notes on the Naga tribes in communication with Assam. 1844.

[Butler, JNIaj. J.] A sketch of As.sam, with some account of the Hill tribes.
London 1S47.

Sketch of the Siitgphos, or the Kakhyens of Burmah; the position of this
tribe as regards Baumo, and the inland trade of the valley of the
Irrawaddy with Yunan and their connection with the north-eastern
frontier of Assam. Calcutta 1847.

Oldham, Th. On the geological structure of part of the Khasi hills, with
observations on the meteorology and ethnology of that district. 4"
Calcutta 1854.

Butler, M.-^j. J. Travels and adventures in Assam during a residence of
fourteen years (1841 — 54). London 1855.

Mc CuLLOCH, M.\j. W. Account of the valley of Munnipore and of the Hill
tribes, with comparative vocabulary. Calcutta 1859 iSelections from
the records of the Govt, of India, Foreign Department, No. X.WIII).

Cooper, T. T. The Mishmee hills, an account of a journey made in an
attempt to penetrate Tibet from Assam, to open new routes for com-
merce. London 1873.

Selection of papers regarding the hill tracts between Assam and Burmah
and on the Upper Brahmaputra. Bengal Secretariat Press 1873.

Hunter, W. W. A statistical account of Assam. 2 vols. London 1879.

Goonabhiram Borooah. Assam Buranji, or. the history of Assam. Including
the history of the ancient kingdom of Kamrup from the earliest times
to the end of 1S75; and containing a brief notice of the castes,
language, religion, commerce, agriculture, arts and social customs of the
people, and the internal government of the province. Calcutta 1876.

Bastlan, A. Volkcrstamme am Brahmaputra und verwandtschaftliche Nach-
barn. Berlin 1883.

194 5- Ethnography.

Ward, Mrs. S. R. A glimpse of Assam. Calcutta 1884.

Way, G. a. Report on the Alchas, the Akha country, and the Akha ex-
pedition. Simla 1884.

Bishop, S. O. Sketches in Assam. Calcutta 1885.

SopPiTT, C. A. A short account of the Kachcha Naga-Empco tribe in the
North Cachar hills, with an outline grammar. Shillong 1885.

SoPPiTT, C. A. An historical and descriptive account of the Kachari tribe
in the North Cachar hills, with specimens of tales and folk-lore.
Shillong 1885; — reprinted, with an introduction by B. C. Stuart
Baker, 1901.

SoppiTT, C. A. A short account of the Kuki-Lushai tribes on the north-
east frontier (districts Cachar, Sylhet, Naga-hills etc., and the North
Cachar hills). Shillong 1887.

Browne, Capt. H. R. The Lushais 1873 to 1889. fol. Shillong 18S9.

Chambers, Capt. O. A. Handbook of the Lushai country. Calcutta 1889.

Davies, A. W. Gazetteer of the North Lushai hills. Shillong 1894.

Anderson, J. D. A collection of Kachari folk-tales ^nd rhymes. [Shillong] 1895.

Carey-, B. S., and H.N. Tuck. The Chin hills: a history of the people,
our dealings with them, their customs and manners, and a gazetteer
of their country. 2 vols. Rangoon 1896.

M.vhendr.anatha Ghoshi. a brief sketch of the religious beliefs of the
Assamese people. Calcutta 1896.

Gurdox, p. R. T. Some Assamese proverbs. Transliterated and trans-
lated into English with explanatory notes. Shillong 1896; — 2"^ ed.
Shillong 1903.

Johnstone, Sir J.\mes. My experiences in Manipur and the Naga hills,
1874—86. London 1896.

Notes on some industries of Assam, 1884 — 95. This is a reprint of: i) Silk
in Assam, by E. Stack, 1884; — 2 1 Cotton in Assam, by H. Z. Darrah,
1885; — 3) Three manufactures of Assam: a) the brass work of the
Morias, b) the gold enamelling of Jorhat, and c) Iron smelting in the
Khasi hills, by H. Z. Darrah, 1S85; — 4) The Eri silk of Assam, by
H. Z. Darrah, 1890; — 5) Brass and copper wares in Assam, by E. A.
Gait, 1894; — 6) Pottery in Assam, by E. A. Gait, 1895. Compiled in
the office of the Secretary to the Chief Commissioner of Assam. 1896.

A list of occasiotial articles relating to Assam and the countries adjoining,
to be found in the journals and books in the Shillong Government Library.
Shillong 1905.

Gait, E. A. A history of Assam. Calcutta 1906.

Gordon, P. R.T. The Khasis. With an introduction by Ch. Lyall. London 1907.

HoDSON, T. C. The Meitheis. With an introduction by Sir Charles J. Lyall.
London 1908.

Stack, E. The Mikirs. From the papers of the late E. Stack. Edited,
arranged, and supplemented by Sir Charles Lyall. London 190S.

Playfair, a. The Garos. With an introduction by Sir J. Bampfylde Fuller.
London 1909.

HoDSON, T. C. The Naga-tribcs of Manipur. London 1911.

Bengal, includinc; Orissa and CHiTTACONn. 195

Bengal, including Orissa and Chittagong.

A rare and most strange relation from Beiigalla in the East Indies, being

one of the greatest kingdoines under the Great Mogull, and of their

lazus, manners and ciistomes, etc. — A briefe relation of the great city

of Jaggarnat. 4" London 1633.
Stirling, A. An account, geographical, statistical, and historical of Orissa

proper, or Cuttack. 4° |Scrampore 1825. J
Buchanan, Fr. A geographical, statistical, and historical description of

the district, or zilla, of Dinajpur, in the province, or soubah, of Bengal.

Calcutta 1S33.
Taylor, J. A sketch of the topography and statistics of Dacca. Calcutta 1840.
Sterling, A. Orissa; its geography, history, religion, and antiquities. Also

history of the Baptist mission by James Peggs. London 1846.
B. KiSHORE Ghose. The history of Poorec, with an account of Juggunnath;

also a succinct description of the southern division of zillah Cuttack.

Calcutta 1S48.
History of the rise and progress of the operations for the suppression of

human sacrifice and female infanticide, in the hill tracts of Orissa.

Calcutta 1854 (Selections from the records of the Govt, of India, No. V).
Sherwill, Mai. J. L. Geographical and statistical report of the district of

Bhaugulpoor. fol. Calcutta 1854.
Sherwill, Capt. \V. S. Geographical and statistical account of the district

Beerbhoom. 4° Calcutta 1855.
Gastrell, J. E. Statistical and geographical report of the Moorshedabad

district. 4° Calcutta i860.
G.\strell, J. E. Statistical and geographical report of the district of

Bancoorah. fol. [Calcutta 1863.]
Thompson, Maj. G. H. A geographical, statistical, and general report on

the district of Hazareebaugh. 4° Calcutta 1864.
Sherwill, M.aj. J. L. Geographical and statistical report of the Dinagepore

district, 1863. 4° Calcutta 1865.
Thompson, Maj. G. H. Notes, geographical, statistical and general, on

that portion of the Lohardagga or Chota Nagpore district, known as

purgunnah Palamow; written during 1862 to 1866. 4° Calcutta 1866.
Man, E. G. Sonthalia and the Sonthals. Calcutta 1S67.
G.\strell, J, E. Geographical and statistical report of the districts of Jessore,

Fureedpore, and Backergungc. 4° Calcutta 1868.
Hunter. W. W. The annals of rural Bengal. 3 vols. London 1868/72 —

I: The annals. (S'^ed. 1872; — 7'h ed. 1897.) — 11—111: Orissa; or,

the vicissitudes of an Indian province under native and British rule.
Lewin, T. H. The hill tracts of Chittagong and the dwellers therein;

with comparative vocabularies of the hill dialects. Calcutta 1S69. —

.A new ed., without the vocabularies: Wild races of South-Eastern

India. London 1870.
Mitchell. ]Mks. M. A missionary's wife among the wild tribes of South

Bengal. Extracts from the journal of Mrs. M. Mitchell. With introduc-
tion and supplement by Dr. G. Smith. Edinburgh 1871.
Westl.\nd. J. A report on the district of Jessore ; its antiquities, its history,

and its commerce. Calcutta 1871; — revised ed. Calcutta 1874.

196 5- Ethnograi'HY.

Dalton, E. T. Descriptive ethnology of Bengal. 4° Calcutta 1872.
Lewin, T. II. Hill proverbs of the inhabitants of the Chittagong tracts.

fol. Calcutta 1873.
Hunter, W. W. A statistical account of Bengal. 20 vols. London 1875 '77.
Rajendrai.ala Mitra. Antiquities of Orissa. 2 vols. fol. Calcutta 1875 80.
Beveridce, H. The district of Bakarganj (eastern Bengal); its history and

statistics. London 1876.
Shib Chunder Bose. The Hindoos as they are. Description of the manners,

customs, and inner life of Hindoo society in Bengal. London-Calcutta

1881; — 2"'' ed., revised and enlarged, 1883.
BuLLORAM MuLLicK. Essays on the Hindu family in Bengal. Calcutta 1882.
Lal Behari Day, Rev. Folk-tales of Bengal. London 1883; — new ed.

London 1899.
Mackenzie, Sir Alex. History of the relations of the government with

the hill tribes of the N.-E. frontier of Bengal. Calcutta 18S4.
Grierson, G. a. Bihar peasant life, being a discursive catalogue of the

surroundings of the people of that province. Calcutta 1885.
BuLLORAM MuLLicK. Home life in Bengal, an account of the everyday

life of a Hindu home at the present day. Calcutta 1885.
Riebeck, E. Die Hiigelstamme von Chittagong. Ergebnisse einer Reise

im Jahre 1882. fol. Berlin 1885.
Christian, J. Behar proverbs, according to their subject matter, and trans-
lation into English with notes, illustrating the social custom, popular

superstition, and every-day life of the people. London 1 891.
Grier.son, G. a. Notes on the district of Gaya. Early history, the geography

of the district, system of land measurement, etc. ; beggars, food of

the poorer classes, their houses, clothes, cooking utensils, ornaments,

marriage and funeral expenses; the cattle of the district. Calcutta 1893.
Oldham, W. B. Some historical and ethnical aspects of the Burdwan

district. With an explanatory index. Calcutta 1894.
Campbell, A. Santal folk tales, translated from the Santali. Pokhuria [1896J.
Lal Behari Day, Rev. Bengal peasant life. (Glossary of Indian terms.)

London 1898.
Abinas Chandra Das. The Vaisya caste. I. The Gandhavaniks of Bengal.

Calcutta 1903.
Bradley-Birt, F. B. Chota-Nagpore : a little-known province of the Empire;

with an introduction by the Earl of Northbrook. London 1903 ; —

2"'' ed., enlarged. London 1910.
Pramatha N\th Mallik. History of the Vaisyas of Bengal. Calcutta 1903.
Hahn, F. Kurukh folk-lore. Collected and transliterated. Calcutta 1905 ;

— (German translation:) Blicke in die Geisteswelt der hcidnischen

Kols. Sammlung von Sagen, Marchen und Liedern der Oraon in

Chota Nagpur. Gutersloh 1906.
District Gazetteers of Bengal, ed. by L. S. S. O'Mallcy. Calcutta 1906 s(iq.
GovIndra N.\th Dutt. The Brahmans and Kayasthas of Bengal. Madras 1906.
Hutchinson, R. H. S. An account of the Chittagong hill-tracts. With 5

appendices. Calcutta 1906.
Hahn, F. Einfiihrung in das Gcbict der Kol-Mission. Gcschichte, Gebrauche,

Religion und Christianisierung der Kols. Gutersloh 1907.

KoNDH. — North -Western Provinces and Oudh. 197

Eastern Bengal District Gazetteers: Chittagong, by L. S. S. O'Malley. 2 vols.

Calcutta 1908.
BoMPAs, C. H. Folklore of the Santal Parganas, translated. London 1909.

The Kondhs.

Mac I'HEKSON, S. Ch. Report upon the Khonds of the districts of Ganjam

and Cuttack. fol. Calcutta 1842; — new ed. fol. Madras 1863.
Campbell, J. A personal narrative of thirteen years' service among the

wild tribes of Khondistan, for the suppression of human sacrifice.

London 1864.
Macpherson, S. Ch. Memorials of service in India . . ., cd. by his brother

William. London 1865.
For a bibliography on the Kondhs, see the article "Kondh" in Thurston's:

Castes and tribes of southern India. Vol. Ill, 412 sqq.

United Provinces of Agra and Oudh.

IsHLREE Dass. Domestic manners and customs of the Hindoos of northern
India, or, more strictly speaking, of the North- West Provinces of India.
Benares i860; — 2"'' ed. Benares 1866.

Moore, W. R. Papers on the subject of Mr. W. R. Moore's investigation
and report regarding female infanticide in the Benares division.
Calcutta 1868.

Sherrixg, M. a. The sacred city of the Hindus: an account of Benares
in ancient and modern times. With an introduction bv F. Hall.
London 1868.

Elliot, H. M. IMemoirs on the history, folk-lore, and distribution of the
races of the North-Western Provinces of India, being an amplified
edition of the supplemental glossary of Indian terms. Revised by
J. Beames. 2 vols. London i86g.

Oldham, W. Historical and statistical memoir of the Gazeepoor district.
2 parts fol. Allahabad 18706.

Atkinson, E. T. Statistical, descriptive, and historical account of the North-
West Provinces of India. 14 vols. Allahabad 1S74 — 84 (I Bundelkhand :
11,111 ^leerut; IV Agra; VRohilkand: \'l Cawnpore, Gorakhpore, and
Basti; VII Farukhabad and Agra; VllI Muttra, Allahabad, Fatehpur;
IX Shahjahanpur, Moradabad, Rampur; X— XII Himalayan Districts;
XIII Azamgarh, Ghazipur, and Ballia; XIV Benares, Mirzapur, Jaunpur).

Growse, F. S. Mathura, a district memoir : history, literature, language
(with glossary), topography, geography, religion, mythology, antiquities,
ethnography, architecture, etc. Printed at the N.-W. Prov. and Oudh
Govt. Press 1874; — 2"'' ed. 4" 1880; — i^ ed., revised and abridged.
4° 1883.

Gazetteer of the provinces of Oudh, published by authority. 3 \<'ls- Luck-
now 1877/8.

Atkinson, E. T. Notes on the history of religion in the Himalaya of the
North-Western Provinces, India. Calcutta 1883.

Growse, F. S. Bulandshahr, or, sketches of an Indian district; social,
historical, and architectural. 4° Benares 1S84.

198 5- Ethnography.

Neskieli), J. C. Brief view of the caste system of the North -Western
Provinces and Oudh, together with an examination of the names and
figures shown in the census report, 1882. foj. Allahabad 1885.

Crooke, W. An ethnographical handbook for the North-Western Provinces
and Oudh. Allahabad 1S90.

Crooke, W. The tribes and castes of the North-Western Provinces and
Oudh. 4 vols. Calcutta 1896.

Crooke, W. The North-Western Provinces of India, their history, ethnology,
and administration. London 1S97.

PiM, A. W. A monograph on woollen-fabrics in the united provinces of
Agra and Oudh. A monograph descriptive of the process of spinning,
warping, weaving, felting namdas, with a sketch of the traditions and
customs connected with wool and woollen fabrics at present, and
conditions and prospects of the industry. 4" 1898.

Dampier, G. R. a monograph on brass and copper-wares of the united
provinces of Agra and Oudh. A monogranh on the origin of vessels
of domestic use, rules concerning the use of brass and copper vessels
in Hindu and Muhammadan scriptures, and popular prejudices con-
cerning them. 40 1899.

District Gazetteer of the United Provinces of Agra and Oudh, ed. by H. R.
Nevill. Allahabad 1903 sqq.

Havell, E. B. Benares, the sacred city. Sketches of Hindu life and religion.
London 1905 (Boston 1905: 1908; — New York 1906); — 2"'' ed. 1911.

Oakley, E. S. Holy Himalaya. Religion, traditions, and scenery of a
Himalayan province (Kumaun and Garhwal). London 1905.

Ganga Datt Upreti. Descriptive Ii.st of the martial castes of the Almora
District, United Provinces. Lucknow 1907.

Panjab.

Gerard, A. Account of Koonawur in the Himalaya. Ed. by G. Lloyd.
London 1841.

Thomas, G. P. Views of Simla: twenty-four lithographs and map, with
descriptions and accounts of the inhabitants, customs, and productions
of the Himalayas, fol. London 1846.

Report showing the relations of the British government zviih the tribes, in-
dependent and dependent, on the jV.- W. Frontier of the Punjab. Calcutta
1856 (Selections from the records of the Govt, of India, No. XII).

Baden-Powell, R. H. Handbook of the economic products of the Punjab:
economic raw products. Roorkee 1S6S.

Griffin, L. H. The Rajas of the Punjab; being the history of the principal
.states in the Punjab and their political relations with the British
government. Lahore 1870; - 2'<J ed. London 1S73.

Harcourt, a. F. p. The Himalayan districts of Kooloo, Lahoul, and ."^piti.
London 1871.

Baden-Powell, B. H. Handbook of the manufactures and arts o( the
Punjab. Lahore 1872.

Mi-HAMMAD Abdul Ghafvr. A complete dictionary of the terms used by
criminal tribes in the Punjab, together with a short history of each
tribe, and the names and places of residence of individual members, etc.
Lahore 1879.

Panjab. — Sikh. — Rajputana. 199

Gazetteer of the Punjaub, i?>9,i—i^. 9 vols. 4» Lahore and Calcutta 1883/4.
Ibbetson, D. Ch. J. Outlines of Panjdb ethnography, being extracts from

the Panjab census report of 1881, treating of religion, language, and

caste. 4" Calcutta 1883.
Fan jab Notes and Queries, a monthly periodical, ed. hy R. C. Temple. 4 vols.

4" Allahabad 1883 '87 (Continued as: North-Indian Notes and Oucries,

by W. Crooke).
Ross, D. The land of the five rivers, and Sindh ; sketches, historical and

descriptive. London 1883.
Temple, R.C. The legends of the Panjab. 3 vols. Bombay-London 18S3 — 1901.
Thorburn, S. S. Musalmans and money-lenders in the Punjab. London 1886.
MuH.^MMAD L.\TIF. Lahore : its history, architectural remains, and anti-
quities, with an account of its modern institutions, inhabitants, their

trade, customs, etc. Lahore 1892.
Wilson, J. Gazetteer of the Shahpur district. Revised ed. Lahore 1897.
Maclagan, E. D. Gazetteer of the Multan district, 1901— 2. Revised ed.

Lahore 1902.
UsBORNE, C. F. Punjabi lyrics and proverbs. Translations in verse and

prose. Lahore 1905 (London 19061.
Dracott, a. E. Simla village tales; or, folk-tales from the Himalayas.

London 1906.
Gazetteer of districts in the Panjab. I904sqq.

The Sikhs.

Prinsep, H. T. Origin of the Sikh power in the Punjab, and political life

of Runjeet Singh, with an account of the present condition, religion,

laws, and customs of the Sikhs; compiled from a report by Capt.

Wm. Murray and from other sources. Calcutta 1834.
M'Gregor, W. L. The history of the Sikhs, containing the lives of the

gooroos, the history of the independent Sirdars, or Missuls, and the

life of the great founder of the Sikh monarchy, Maharajah Runjeet

Singh. 2 vols. London 1846; — a-"* ed. 1S47.
Cunningham, J. D. A history of the Sikhs, their country, and the battles

of the Sutlej. London 1849; — 2"^ ed. London 1853; — reprinted.

Calcutta 1904.
Honigberger, J. M. Friichte aus dem Morgenlande. Wien 1851.
Trumpp, E. The Adi-Granth, or, the holy scriptures of the Sikhs. Translated

from the original Gurmukhi, with introductory essays. 4° London 1877.
Trumpp, E. Die Religion der Sikhs. Nach den Quellen dargestellt.

Leipzig 1 88 1.
Macauliffe, M. A. The Sikh religion, its gurus, sacred writings, and

authors. 6 vols. Oxford 1909.

Rajputana.

Tod, J. Annals and antiquities of Rajasthan, or the central and western
Rajpoot states of India. 2 vols. 4° London-Calcutta 182932; —
2"'! ed. 2 vols. 4° Calcutta 1877 79; — reprinted, 2 vols. Madras

200 5- Ethnography.

1873: — do., Madras 1880; — other edd.: 2 vols. 8" Calcutta 1894;
— do., Calcutta 189799; — do., Calcutta 1902—3.

Bon.EAC, A. H. E. Personal narrative of a tour through the western states
of Rajwara, in 1835, comprising Beekaner, Jesulmcr, and J(>dh|)Oor,
with the passage of the Great Desert, and a brief visit to the Indus
and to Buhawalpur. 4" Calcutta 1837.

Dixon, C.J. Sketch of Mairwara: giving a brief account of the origin and
habits of the Mairs; their subjugation by a British force; their civili-
sation, and conversion into an industrious peasantry. 4" London 1850.

Walter, Maj. C. K. M. A gazetteer of the Bhurtpore state. Agra 1868.

PowLETT, Maj. P. W. Gazetteer of the Bikanir state. Calcutta 1874.

PowLETT, M.\j. P. VV. Gazette of the Karauli state. Calcutta 1874.

Latouche, Sir J. J. Digges. Gazetteer of Ajmere-Merwara in Rajputana.
Calcutta 1875.

PowxETT, M.^j. P. W. Gazetteer of Uhvar. London 1878.

The Rajputana gazetteer, compiled by La Touche, C. A. Bayley, H. B. Abbott,
C. K. M.Walter, C. E. Yate, etc. 3 vols. Calcutta and Simla 1S79 So.

J.\coB, Col. S. S., and T. H. He.ndlev. Jeypore enamels. 4" London 1S86.

Hendlev, T. H. Ulwar and its art treasures, fol. London 1 888.

Adams, A. The western Rajputana states, a medico-topographical and
general account of Marwar, Sirohi, Jaisalmir. London 1899.

The Rajputana district gazetteers ;\'o\.\: Ajmer-Merwara, text and statistical
tables. Compiled by C. C. Watson. 2 vols. Ajmer 1904.

Sind.

PoTTiNGER, H. Travels in Beloochistan and Sinde, with geographical and
historical account of those countries. 4" London 1816: —

Reisen durch Beloochistan und Sinde, nebst geogr. u. hist. Nachrichten
iiber diese Lander. Aus d. Engl. Weimar 1817; —

Voyages dans le Bcloutchistan et Ic .Sindh : traduit par Evrics. 2 tomes
Paris 1818.

Burses, J. A narrative of a visit to the court of Sinde, a sketch of the
history of Cutch, and some remarks on the medical topography of
Bhooj. London 1 831; — (also: Edinburgh 1 831); — new ed., with
appendices by Sir H. Pottinger and Sir Alex. Burnes. Edinburgh 1839.

Postans, T. Personal observations on Sindh; the manners and customs

of its inhaliitants, and its productive capabilities, with a sketch of its

history, etc. London 1843.
Burton, Capt. R. F. Sindh, and the races that inhabit the valley of the

Indus; with notices of the topography and history of the province.

London 1851.
Hughes, A. W. A gazetteer of the province <>f Sind. London 1874: —

2"'' ed. London 1876.
RocHiRAM Gajumal. Handbook of Sindhi proverbs with English renderings

and c(iuivalent sayings. Karachi 1S95.

Bombay-Presidency.

Bombay-Presidency.

Steele, A. Summary of the law and custom of Hindoo castes within the
Dekhun provinces subject to the presidency of Bombay, chiefly affecting
civil suits, fol. Bombay 1827; — new ed. London 1868.

Mackintosh, A. An account of the origin and present condition of the
tribe of Ramoossies, including the life of the chief Oomiah Naik.
Bombay 1833.

PosTANs, Mrs. M. Cutch, or, random sketches taken during a residence ;
interspersed with legends and traditions. London 1839.

Gr.\ham, Capt. D. C. a brief sketch of the Bheel tribes, inhabiting the
province of Khandesh. [Bombay 1843.]

PosTANs, Mrs. M. Travels and adventures in Western India; with the
domestic manners and customs of its inhabitants. 2 vols. London 1847.

Briggs, H. G. The cities of Gujarashtra, their topography and history. 4°
Bombay 1849.

MoDV, C. R. An essay on female infanticide. To which the prize offered
by the Bombay Govt, for the second best essay against female in-
fanticide among the Jadayas and other Rajpoot tribes of Guzerat was
awarded. Bombay 1849.

Hervev, Ch. Reports on the Kaikarec, Dacoitee, and predatory tribes
infesting Bombay presidency, 1849 — 52. fol. Bombay 1853.

Gilder, C. On Hinduism, and the principal Hindu deities worshipped in
the Bombay presidency. Bombay 1855.

Forbes, A. K. Ras Mala; or, Hindoo annals of the |)rovince of Goozerat,
in Western India. 2 vols. London 1856: — new ed., with an intro-
duction by Maj. J. W. Watson, and a memoir by A. K. Nairne. London-
Calcutta 1S78.

Johnson, W. The oriental races and tribes, residents and visitors of Bombay.
2 vols. 4° London 1865/66. (Vol. I: Gujarat. Kutch, and Kathiawar:
vol. II: Maharashtra.)

[Karsandas Muljl] History of the sect of Maharajas or Vallabhacharyas
in Western India. London 1865.

Stokes, H. J. An historical account of the Beigaum district in the Bombay
presidency. Bombay 1870.

KiEs, G. Geographical and historical description of the southern Maratha
country and North-Kanara, comjirising the four collectorates of Kaladgi,
Beigaum, Dharwar and North Kanara. fol. Bombay 1876.

Wilson, Rev. J. Aboriginal tribes of the Bombay presidency. A fragment.
Bombay 1876.

The Gazetteer of the Bombay presidency, by Sir J. M. Campbell. Vols. I — XXVII.
Bombay 1877 — 1907.

Gunthorpe, Maj. E. J. Notes on criminal tribes residing in or frequenting the
Bombay presidency, Berar, and the Central Provinces. Bombay 1882.

M.\LABARi, B. M. Gujarat and the Gujaratis : pictures of men and manners
taken from life. London 1882: — 2"<i ed. Bombay 1884: — 3'' ed.
Bombay 1889.

GuPTE, B. A. Catalogue of Bombay contributions to the economic court
of the Calcutta International Exhibition of 1883 '84. Part I: Ethnology.
Calcutta 1884.

5- Ethnography.

Scott, Col. W. A .short account of the Gimdal state. Bombay 1886.
Douglas, J. Bombay and Western India, a series i>f stray papers. 2 vols.

London 1893.
Damldhai Davabhai Mehta. Gujerati proverbs, collected and alphab.

arranged with an essay on proverbs. Ahmedabad 1894.
The ethnographical survey of Bombay, Monographs. Bombay 1903 sqq.

(Govt. Central Press.)
Kincaid, C. a. The outlaws of Kathiawar, and other studies. Bombay 1905.
Enthoven, R. E. Notes for a lecture on the tribes and castes of Bombay. 1907.
Kennedy, M. Notes on criminal classes in the Bombay presidency.

Bombay 1908.

Edwarixs, S. M. Gazetteer of Bombay City and Island. 3 vols. Bombay
1909 — 10.

The Parsis.

Anquetil du Person. Le Zend Avesta. 3 tomes; tome II: Usages civils

et religieux des Parsees. 4° Paris 1 771.
Briggs, H. G. The Parsis, or, modern Zerdusthians. Edinburgh 1852.
DosABHAi Framjee Karaka. The Parsees : their history, manners, customs,

and religion. London 1858.
Haug, M. Essays on the sacred language, writings, and religion of the

Parsis. Bombay 1862; — 2"d ed. London 1S78: — 3^ ed. London 1884.
Dosabhai Framjee Karaka. History of the Parsis, including their manners,

customs, religion and present position. 2 vols. London 18S4.
Menant, D. Les Parsis, histoire des communautcs zoroastriennes de I'lnde.

I'' partie. Paris 1898 (Annales du Musce Guimet, bibliothcque d'etudes,

tome VIIi.
Modi, J. J. Marriage customs among the Parsees, their comparison with

similar customs of other nations. Bombay 1900.

The Marathas.

Broughton, T. D. Letters written in a Mahratta camp during the year 1809,
descriptive of the character, manners, domestic habits, and religious
ceremonies of the Mahrattas. 4" London 1813: — new ed., with an
introduction by M. E. Grant Duff. Westminster 1892 : —
Wanderungen unter den Mahratten (1809). Aus dem Englischen.
Leipzig 1 8 14; — Les Marattas ou moeurs, usages et costumes de ce
peuple: traduit de I'anglais par M. Breton. 2 tomes 12° Paris 1817.

Duff, J. G. A history of the Mahrattas. 3 vols. London 1826: — reprinted,
3 vols. Bombay 1863: — 3d ed. I vol. Bombay 1873.

Taylor, Col. Meadow.s. Tara, a Mahratta tale. 3 vols. London 1863: —
new ed. 1874.

[HocKLKY, W. B.] Pandurang Hari, or, memoirs of a Hindoo. With an
introductory preface by Sir H. B. E. Frere. 2 vols. 12" London 1873;
— new ed. i vol. S" London 1883.

Manwaring, A. Marathi proverbs, Marathi and English. Oxford 1S99.

Ganesh Narayan De.shpande. Dictionary of Marathi proverbs. Poona 19CX).

Central India, Central-Provinces, etc. — Southern India, general. 203
Central India, Central-Provinces, Berar and Haidarabad.

Leckie, D. R. Journal of a route to Nagpore, by the way of Cuttac,

Burrosumber, and the Southern Kunjare Ghaut in the year 1790, with

an account of Nagpore: and a journal from that place to Benares,

by the Soohagee pass. 4" London 1800.
Malcolm, Sir John. A memoir of Central India, including Malwa, and

adjoining provinces. 2 vols. London 1823: — 2"'' ed. London 1824: —

3<* ed. London 1832; — reprinted from the 3'' cd., 2 vols. London-
Calcutta 1S80.
Malcolm, Sir John. Rei)ort on the i)rovinccs of Malwa and adjoining

districts. 2 vols. 4° Calcutta 1828.
PoGSON, W. R. A history of the Boondclas. 4" Calcutta 1828.
Hough, W. Bhopal principality, Central India: a brief history, 1700 — 1845.

Calcutta 1845.
HisLOP, Rev. St. Papers relating to the aboriginal tribes of the Central

Provinces, left in Mss. by the late Rev. S. Hislop. Edited, with notes

and preface, by R. Temple. [Nagpore] 1866.
Gr.\nt, Ch. Gazetteer of the Central Provinces of India. 2"''ed. Nagpur 1870.
Lyall, a. C. Gazetteer for the Haidarabad assigned districts, commonly

called Berar, 1870. Bombay 1870.
For.svth, J. The highlands of Central India: notes on their forests and

wild tribes, etc. London 1871: — • 2"^ ed. London 1872; — new ed.

London 1889: — new ed. London 1899.
S.wiD HusAiN AND C. Wilmott. Historical and descriptive sketch of His

Highness the Nizam's dominions. 2 vols. Bombay 1883^4.
Bloomfield, Lieut. -Col. A. Notes on the Baigas of the Central Provinces

of India. Nagpur [1885].
Fuller, J. B. Review of the progress of the Central Provinces during

the past 30 years and of the present and past condition of the people.

fol. Nagpur 1892.
District Gazetteers, Central Provinces. Vol. A, descriptive. Bombay and

Allahabad i9o6sqq.
Gordon, E. M. Indian folk-tales. Being side-lights on village life in Bilaspore,

Central Provinces. London 1908.
Luard, C. E. a bibliography of the literature dealing with the Central

India Agency, to which is added a series of chronological tables.

London 1908.
Luard, C. E., and Pandit K.^shin.^th Krishna Lele. The Paramaras of

Dhar and Malwa. Bombay 1908.
Ethnographical survey of the Central India Agency, ed. by Capt. C. E. Luard.

Lucknow 1909.

Southern India, general.

Baldaeus, Ph. Naauwkeurigc beschrijvinge van Malabar en Choromandel . .
en het eyland Ceylon, nevens dc afgoderije der Oost-lndische Heij-
denen etc. fol. Amsterdam 1672; —
Wahrhaftige ausfiihrliche Beschreibung der beriihmten Ost-Indischen
Kiisten Malabar und Coromandel, als auch der Insel Zeylon etc. fol.
Amsterdam 1672 : —

204 5- Ethnography.

A description of yc East India coasts of Malabar and Coromandel, with
their adjacent Kingdoms and provinces; and of the empire of Ceylon:
and of the idolatry of the pagans in the Eqst Indies, fol. London
'703; — another ed. London 1747.

Buchanan, F. A journey from Madras through the countries of Mysore,
Canara, and Malabar, |)erformcd under the orders of the Marquis
Wclleslcy, for the express ]>urpose of investigating the state of agri-
culture, arts, and commerce: the religion, manners, and customs: the
history, natural and civil, and antiquities of the dominions of the Rajah of
Mysore. 3 vols. 4" London 1807: — new ed. 2 vols. 8» Madras 1870.

L'Inde fraiifaise, ou collection de dessins re])rcs. les divinites, temples,
meubles, ornements, armes, ustensiles des peuplcs hindous cjui habitent
les possessions franqaises de I'lnde, et en general la cote de Coro-
mandel et de Malabar. Publiee par M. J. J. Chabrelie, avcc un texte
explicant par M. E. Burnouf et M. E. Jacquet. 2 tomes fol. Paris
1827 35: — 2"= ed., dessinee et publiee par Geringer et Chabrelie,
et accompagnce d'un tcxte explicatif rcdige par M. E. Burnouf et
M. E. Jacquet. 2 tomes fol. Paris (ca. 1870).

Ramasvami, C. V. Descriptive and historical sketches of cities and places
in the Dekkan. With a brief description of the southern peninsula,
and a succinct history of its ancient rulers. Calcutta 1828.

Wn.soN, H. H. Mackenzie Collection. A descriptive catalogue of the
oriental manuscripts, and other articles illustrative of the literature,
history, statistics and antiquities of the south of India: collected by
the late Lieut.-Col. C.Mackenzie. 2 vols. Calcutta 1828: — 2"'' ed.
Madras 1882.

ZiECENBAi.G, B. Genealogie der Malabarischen GiJtter. Erster, ungean-
derter, notdiirftig erweiterter Abdruck, besorgt durch Dr. W. Germann.
Madras 1867: —
English translation, with., additions, by the Rev. G. J. Metzger. Madras 1869.

Frere, Mi.ss M. Old Deccan days, or, Hindoo fairy legends, current in
southern India. London 1868: — 3'' ed., revised. London 1881: —
S''' ed. London 1898.

Kearnes, Rev. J. F. Kalayan'a Shat'anku, or, the marriage ceremonies of
the Hindus of South India, together with a description of Karumman-
tharum, or, the funeral ceremonies. Madras 1868.

Govek, Ch. E. The folk-songs of southern India. Essay and translations.
Madras 1871: — London 1872.

Baierlein, E. R. Nach und aus Indien. Leipzig 1873: —

The land of the Tamulians and its missions. Transl. from the German-
by J. D. R. Grihble. Madras 1875.

Sewei.i,, R. a sketch of the dynasties of southern India. Compiled under
the orders of Government. 4" Madras 1883.

Bhandarkar, R. G. Early history of the Dekkan down to the Mahomcdan
conquest. Bombay 1884: — 2">' ed. Bombay 1895.

Rice, H. Native life in South India, being sketches of the social and
religious characteristics of the Hindus. London 1889.

KiNCScoTE, H., AND N. .Sastri. Tales of the sun, or folklore of southern
India. London i8go.

SoUIHERN-IXDIA, GENERAL. — MadRAS-PrESIDENCY. 20$

Day, Cai't. C. 1\. The music and musical instruments of soutiiern India and

the Dcccan. With an introduction by A. J. Hipkins. 4'> London 1891.
Bulletin of the Madras Government Museum. Ed. by E. Thurston.

Madras 1894 sciq.
Gribble, J. D. B. A history of the Deccan. Vol. I. London 1896.
Rice, S. P. Occasional essays on native South-Indian life. London (New

York) 1901.
MiMAN'DE, P. Le Mamoul. Paris 1904. [Religious and social customs of

S. India]
Thurston, E. Ethnographic notes in southern India. Madras 1906: —

new ed. Madras 1908.
Ml'ller, D"^ H. Untersuchungen iiber die Gcschichte der polvandrischen

Eheformen in Siidindien. (Teil I, Kap. 4 iiiul Tcil 11, Kap. L Berlin

1909 (Dissertation, Bonn).
Thurston, E., and K. Rangachari. Castes and tribes of southern India.

7 vols. Madras 1909 10.

Madras-Presidency.

Roger, A. Le theatre de I'idolatrie, ou la porte ouverte, pour parvenir
a la cognoissance du paganisme cache, et la vraye representation de
la vie, des moeurs, de la religion et du service divin des Bramines,
qui demeurent sur les costes de Choromandel, ct aux pays circon-
voisins. 4° Amsterdam 1670.

Le Gentil, G. J. H. J. B. Voyage dans les mers de I'lnde, fait par Tordre du
roi, a I'occasion du passage de Venus, etc. 2 tomes 4° Paris I779'8i;
Tome I: Coutumes et usages des Indiens de la cote de Coromandel ;
de la religion des Indiens tamoults, etc.

Haafner, J. Reize in eenen palanquin, of lotgevallen en merkwaardige

aanteekeningen op eene reize langs de Kusten Orixa en Choromandel.

Amsterdam 1808 (Haarlem 1809); — 2" druk Amsterdam 1826: —

Voyages dans la peninsule occidentale de I'lndc ct dans I'lle de Ceilan.

Traduit du HoUandais par M.J. 2 tomes Paris iSii.

Caldwell, R. The Tinnevelly Shanars, their religion, moral condition,
and characteristics. Madras 1849.

Hamilton, Gen. D. Report on the Pulni mountains. [i86i.|

Wheeler, J. T. ^Madras in the olden time, being a history of the presi-
dency, from the first foundation of Fort St. George. 3 vols. Madras
1861 : — new ed. i vol. Madras 1882.

Hamilton, Gen. D. Report on the Shevaroy hills. Madras 1862.

Hamilton, Gen. D. Sketches of the Shevaroy hills, fol. London 1865.

Hamilton, Gen. D. Report on the high ranges of the Annamullay mountains.
Madras 1866.

Carr, M. W. Collection of Telugu proverbs, together with some Sanscrit pro-
verbs. With a translation, illustr. and explanation. 2 parts Madras 1868.

Nelson, J. H. The JNIadura country, a manual. Madras 1868.

Carmichael, D. F. Manual of the district of Vizagapatam in the pres. of
Madras. Madras 1869.

Descriptive and historical papers relating to the Seven Pagodas on the Coro-
mandel coast, by W. Chambers, J. Goldingham, B. G. Babington, Rev.

2o6 5. Ethnography.

G. W. Mahon, Uexit J. Braddock, Rev. W. Taylor, Sir Walter Elliot,

and C. Gubbins. Ed. by Capt. M. W. Carr. Madras 1869.
Kelsall, J. Manual of the Bellary district. Madus 1872.
BoswELL, J. A. C. A manual of the Nellore district in the presidency of

Madras. Madras 1873.
HiCKEY, VV. The Tanjorc Mahratta principality in southern India: the

land of the Chola, the Eden of the south. Madras 1873: — 2'"' cd.

Madras 1874.
Percival, p. Tamil proverbs, with their English translation. Madras 1874:

— new ed. 1877.
Gribble, J. D. B. A manual of the district of Cuddapah, in the presidency

of Madras. Madras 1875.
Thomas, E. C. G. Vizagapatam, the port for the Central Provinces, fol.

Madras 1877.
Garstin, J. H. A manual of the South-Arcot district. Madras 1878.
Moore, L. A manual of the Trichinopoly district in the presidency ol

Madras. Madras 1878.
Morri.s, H. a descriptive and historical account of the Godavery district

in the presidency of Madras. London 1878.
Crole, C. S. The Chingleput, late Madras district: a manual. Madras 1879.
Jensen, H. Classified collection of Tamil proverbs. Madras 1879: — new ed.

London 1897.
Stuart, A. J. A manual of the Tinnevelly district in the presidency of

Madras. Madras 1879.
Caldwell, R. A political and general history of the district of Tinnevelly,

in the presidency of Madras, from the earliest period to its cession

to the English government in A. D. iSoi. Madras 1881.
Co.x, A. F. A manual of the North-Arcot district in the IMadras-Presidency.

Madras 1881 ; — 2"' ed., revised by H. A. Stuart. 2 vols. Madras 1894/5.
Maltby, T. J. The Ganjam district manual: ed. by G.D. Leman. Madras 1882.
Le Fanl", H. a manual of the Salem district in the presidency of Madras.

2 vols. Madras 1883.
Mackenzie, G. A manual of the Kistna district in the presidency of Madras.

Madras 1883.
Row, T. Venka.saml A manual of the district of Tanjore in the Madras-
Presidency. Madras 1S83.
Narahari Gopalakrishnama Chetty. A manual of the Kurnool district in

the presidency of Madras. Madras 1 886.
Nicholson, F. A. Manual of the Coimbatore district in the presidency of

Madras. 2 vols. Madras 1887, 1898 (Vol. II revised by H. A. Stuart).
Wilson, F. H. The Shevaroys ; a story of their past, present, and prospec-
tive future. Madras 1888.
Benson, Ch. An account of the Kurnool district, based on an analysis of

statistical information relating thereto, and on personal observations.

fol. Madras 1S89.
MuLLALY, F. S. Notes on the criminal classes of the Madras-Presidency.

Madras 1892.
Lazarus, J. A dictionary of Tamil proverbs, with an introduction, and

hints in English on their meaning and application. Madras 1894.

Madras-Presidency. — The Malabar Coast. 207

Benson, Ch. A statistical atlas of the Madras-Presidency: compiled from

existing records, fol. Madras 1895.
Brown, C. P. Essay on the creed, customs, and literatxire of the Jangams,

Madras-Bangalore 1897.
Sewei.l, R. a forgotten empire ( Vijayanagar): a contribution to the history

of India. London 1900.
FooTE, R. B. Catalogue of the prehistoric antiquities, Government Museum,

Madras. Madras 1901.
Gazetteer of districts of the Aladras-Presideiicy. By W. Francis and F. R.

Hemingway. Madras 1904 sqq.
SuBRAMiAH Pantulu, G. R. Folk-lorc of the Telugus. A collection of 42

highly amusing and instructive tales. Madras 1905.
Papa Rao Naidl". The criminal tribes of India : No. 2 : The history of

Korawars, Erukalas, or Kaikaries. Madras 1906.
The Baramahal records. Sect. I : Management : Sect. Ill : Inhabitants.

Madras 1907.
Martin, A. F. Ethnographic notes on the Maduvars. 2"<* ed. Madras 1907.
Wheeler, J. T. Handbook to the Madras records. Being a report on the

public records preserved in the Madras Government Office, previous

to 1834: with chronological annals of the Madras-Presidency. 2"'' ed.

Madras 1907.
Papers relating to the history of the BanganapaUe state. Madras 1908.

(Selections from the records of the Madras Govt.)
A statistical atlas of the Madras-Presidency. Madras 1908.

The Malabar Coast.

Barbosa, Dl'arte. a description of the coasts of East Africa and Malabar
in the beginning of the le''' century-. Translated from an early Spanish
manuscript in the Barcelona library, with notes and preface by the
Hon. Henry E. J. Stanley. London 1866 (Hakluyt-Society, vol. 351.

Phillips, J. T. An account of the religion, manners, and the learning of
the people of IMalabar, from the letters of the Danish missionaries.
12° London 17 17.

VisscHER, J. C. Mallabaarse brieven, behelzende eene naukeurige be-
schrijving van de Kust van Mallabaar, den aardt des landts, de
zeden en gewoontens der inwoneren. Uitgegeven door C. T. Visscher.
Leeuwarden 1743: —
Letters from Malabar . . . Now first transl. from the original Dutch. To
which is added an account of Travancore, and Era Bartolomeo's
travels in that country, by Major H. Drury. Madras 1862.

Apetz, H. Descriptio terrae Malabar. Ex arabico Ebn Batutae itinerario
edita, interpretatione et annotationibus instructa. Jena 1 8 19.

Tohftit-ul-Mujahideen (by Sheikh Zcen-ud-deen), an historical work in the
Arabic language. Translated into English by Lieut. M. J. Rowlandson.
London 1S33: — another ed. of the same work: Zinadim Benali
Benamede, Historia dos Portugueses no Malabar. Ed. Dav. Lopes.
Lisboa 1898.

Lawson, Ch. a. British and native Cochin. 2"'' ed. London 1861.

2o8 5. Ethnography.

Day, F. The land of the Permauls, or Cochin : its past and its present.

Madras 1863.
Phimi'os, Rev. E. The Syrian christians of Malaljar, otherwise called the

christians of S. Thomas. Ed. by the Rev. G. B. Howard. Oxford 1869.
M.\TF,EK, Rev. S. The land of charity: account of Travancorc and its people,

with special reference to missionary labour. London 1 87 1.
.1 thousand Canarese proverbs. Mangalore 1874.
P. Shlncoony Menon. A history of Travancore from the earliest times.

Madras 1878.
\Vii;ram, H. Malabar law and custom. Madras 1882; — 2"<i ed., by L. Moore,

Madras 1900: — 3'' ed. Madras 1905.
Mateer, Rev. S. Native life in Travancore. London 1883.
Miley, S. Canara past and present. Madras 1884.
Logan, W. A manual of the district of Malabar. 3 vols. Madras 1887 — 91;

— 2"'' ed. Madras 1906. '

Ward, B. S. Memoir of the survey of Travancorc and Cochin, 1816 — 1820.

fol. Madras 1891. (Selections from the records of the Madras Govt.)
Rae, Rev. G. M. The Syrian church in India. Edinburgh-London 1892.
BuRNELL, A. C. The devil worship of the Tuluvas. From the papers of

the late A. C. Burnell. Preface by Major R. C. Temple. 4" Bombay

1894. (Reprinted from the "Indian Antiquary".)
Sturrock, J., AND H. A. Stuart. A manual of the district of South-Canara.

2 vols. Madras 1894/5.
GoPAL Panikkar, T. K. Malabar and its folk. A systematic description of

the social customs and institutions of Malabar. With an introduction

by the Rev. F. W. Kcllett. Madras 1900: — 2"'' ed., revised and

enlarged. Madras 1904.
Fawcett, F. Nayars of Malabar. Madras 1901.
Mackenzie, G. T. Christianity in Travancore. Trivandrum 1901.
Shankunny, p. Folklore in Malabar. Calicut 1902.
Ethnological survey of Cochht, monographs. Madras 1905 sqq.
Medlycott, a. E. India and the apostle Thomas, an inquiry. With a

critical analysis of the Acta Thomae. London 1905.
Malabar district gazetteer. By C. A. Innes and F. B. Evans. Vol. I. Madras 1908.
Richards, Rev. W. J. The Indian christians of Saint Thomas, otherwise

called the Syrian christians of Malabar. London 1908.
Ananta Krishna Jyer. The Cochin tribes and castes. With introduction

by A. li. Keane. Vol. I. Calcutta 1910.
Gopalan Nair, C. Malabar Series: Wynad, its peoples and traditions.

Madras 191 1.

Goa.

Burton, Capt. R. F. Goa and the Blue Mountains, or, six months of sick

leave. London 1851.
Fonseca, J. N. DA. An historical and archa:ological sketch of the city of

Goa, preceded by a short statistical account of the territory of Goa.

Bombay 1878.
Contzen, L. Goa im Wandcl dcr Jahrhundcrte. Beitrag zur portugiesischen

Kolonialgeschichte. Berlin 1902.

Mysore and Coorg. — The Nilgiri Hills. 209

Mysore and Coorg.

MiCHAUD, J. Histoire des progrf^s et dc la chiitc dc I'cmpirc dc Mysore,

sous les regnes d'Hyder-Aly ct Tippoo-Saib. 2 tomes Paris 180 1—9.
WiLKS, M. Historical sketches of the south of India, in an attempt to

trace the history of Mysoor: from the origin of the Hindoo government

of that state, to the extinction of the ]NK)hammedan dynasty in 1799.

Founded chiefly on Indian authorities. 3 vols. 4" London 1 8 10— 17;

— 2"^ ed. 2 vols. 8° Madras 1869.
Arthur, Rev. W. A mission to the Mysore ; with scenes and facts illu-
strative of India, its people, and its religion. London 1847.
M5GLING, Rev. H. Coorg memoirs: an account of Coorg and of the Coorg

mission. Bangalore 1855.
MoGLiNG, H., AND Th. Weitbrecht. Das Kurgland uiid die evangelische

Mission in Kurg. Basel 1866.
RicHTER, G. Manual of Coorg : a gazetteer of the natural features of the

country, and the social and political condition of its inhabitants.

Mangalore 1870.
FouLKES, Rev. T. The legends of the shrine of Harihara, in the province

of Mysore. Translated from the Sanskrit. Madras 1876.
Malleson, Col. G. B. Seringapatam, past and present : a monograph.

Madras 1876.
Rice, B. L. Mysore and Coorg. A gazetteer compiled for the government

of India. 3 vols. Bangalore 1876—8: — Mysore gazetteer. Revised cd.

2 vols. London 1897.
RiCHTER, G. Ethnographical compendium on the castes and tribes found

in the province of Coorg: with a short description of those peculiar

to Coorg. fol. Bangalore 1887.
Harris, L. T. Coorg district gazetteer. Vol. B. Mercara 1906.

The Nilgiris.

Hough, J. Letters on the climate, inhabitants, productions, etc., oftheNeil-

gherries, or Blue Mountains of Coimbatoor, South India. London 1829.
Harkness, Capt. H. a description of a singular aboriginal race inhabiting

the summit of the Neilgherry hills, or Blue Mountains of Coimbator,

in the southern peninsula of India. London 1832.
Baikie, R. Observations on the Neilgherries, including an account of their

topography, climate, soil, and productions, and of the effects of the

climate on the European constitution. Ed. by W. H. Smoult. Calcutta

1834: — 2"'! ed. Calcutta 1857.
Jervis, H. Narrative of a journey to the falls of the Cavery, with an

historical and descriptive account of the Neilgherry hills. London 1834.
Metz, J. F. Die Volksstamme der Nilagiris. Basel 1857.
Shortt, D' J. An account of the tribes on the Neilgherries, by J. Shortt:

and a geographical and statistical memoir of the Neilgherry mountains,

by the late Col. Ouchterlony. Madras 1868.
King, W. Ross. The aboriginal tribes of the Nilgiri hills. A paper read

before the Anthropological Society. London 1870.
Breeks, J. \V. Account of the primitive tribes and monuments of the

Nilagiris. 4° London 1873.

Indo-Aryan Research. II. 5. '*

2IO 5- Ethnography.

Marshall, W. E. A phrenologist amongst the Todas, or, the study of a
primitive tribe in South India; with grammar of the Tuda. (Appendix:
A brief outline of the language, by Rev. G. U. Pope.) London 1873.

Grigg, H. B. a manual of the Nilagiri district in the Madras presidency.
Madras 1880.

Francis, W. The Nilgiris (Madras district gazetteer). 2 vols. Madras 1905— 8.

Rivers, W. H. R. The Todas of the Nijgiri hills. London 1906.

Ceylon.

Knox, R. An historical relation of the island of Ceylon in the East Indies, etc.
fol. London 1681: — republished, with an introductory preface and
notes, by W.M. Harvard. London 1821 ; — reprinted. 8° London 1910: —
(Ceylanische Reise-Beschreibung oder historische Erzehlung von der in
Ost-Indien gelegenen Inscl Ceylon etc. 4* Leipzig 1680; — T Eyland
Ceylon in syn binnenste, of 't Koningrijck Candy; geopent en nauw-
keuriger dan oyt te voren ontdeckt. Vcrtaald door S. de Vries. 4°
Utrecht 1692; — Relation du voyage de I'isle de Ceylan . . . les
moeurs, les coutumes, et la religion de ses habitans etc. 2 tomes
Amsterdam 1693.)

Percival, R. An account of the island of Ceylon : its history, geography,
natural history, with the manners and customs of its various inhabitants.
London 1S03: — 2"d ed. London 1805: — ed.in 2 vols. London 1833: —
(Beschreibung von der Insel Ceylon etc. Aus dem Englischen mit
Zusatz iiber die Perlenfischcrey iibersetzt von J. A. Bergk. Leipzig
1803: — Aus dem Englischen iibersetzt von Ehrmann, Weimar 1804
(Sprengels Bibl. d. Reisebeschr. Bd. 11): — Voyage a I'llc de Ceylan,
fait de 1797 a 1800; traduit par F. P. Henry. 2 tomes Paris 1803).

Haakner, J. Reize te voet door het eiland Ceilon. Amsterdam iSio: —
2'<= druk, Amsterdam 1826: — (Fufireise durch die Insel Ceilon: nach
dem Hnllandischen frei bearbeitct. Magdeburg 1 8 16: — Travels on foot
through the island of Ceylon. Translated from the Dutch. London 1 821).

Philalethes. The history of Ceylon to 1815, with characteristic details of
the religion, laws, and manners of the people, and a collection of
their moral maxims and ancient proverbs. To which is subjoined,
Robert Knox's historical relation of the island, with an account of his
captivity. 4° London 181 7.

Daw. An account of the interior of Ceylon, and of its inhabitants, with
travels in that island. 4° London 1821.

Gauttier, E. Ceylan, ou recherches sur I'histoire, la littcraturc, les mceurs
ct les usages des Chingulais. 12° Paris 1823.

Saram, A. DE. A description of castes in the island of Ceylon, their trades
and their services to government ... To which is appended a des-
cription of the dress of native headmen, according to their different
castes, copied from an old Ceylon Almanac dated 181 1. 4" Galle 1823.

Upham, E. The history and doctrine of Budhism, with notices of the
Kappooism, or demon-worship, and of the Bali, or planetary incan-
tations of Ceylon, fol. London 1829.

VaUiiH Naf/amiawii, a Cingalese poem, descriptive of the Ceylon system of
demonology: to which is appended the practices of a capua,'or devil
priest, and Kolan Nattannawa. Translated by J. Callaway. London 1S29.

Ceylon. 2 i i

Chitty, S. C. Ceylon gazetteer: accurate account of the districts, pro-
vinces, etc., of the island of Ceylon : together with sketches of the
manners, customs, etc., of its various inhabitants. Ceylon 1834.

yournal of the Ceylon Branch, Royal Asiatic Society. Colombo 1845 sqq.

SiRR, H. Ch. Ceylon and the Cingalese, their history, government, religion,
produce and capabilities. 2 vols. London 1850.

Tennent, Sir J. Emerson. Christianity in Ceylon: its introduction and
progress under the Portuguese, the Dutch, the British, and American
missions. With an historical sketch of the brahmanical and buddhist
superstitions. London 1850.

Tennent, Sir J. Emerson. Ceylon, an account of the island, physical,
historical and topographical: with notices of its natural history, anti-
quities, and productions. 2 vols. London 1859: — 2"'', j") cdd. 1859:
— 4''^, 5'h edd. i860.

/fihdsa, or a collection of useful information concerning the natives of
Ceylon, as recorded in ancient histories: compiled by Weligama Sri
Suwangaia Terunnanse. Published by A. Dias. Colombo 1876.

The Taprobanian; a Dravidian Journal of oriental studies in and around
Ceylon. Ed. by H. Nevill. Vol. I— III. 4° Bombay-London 1885 '87.

Sarasin, p. UNO F. Ergebnisse naturwissenschaftlicher Forschungen auf
Ceylon, Band 3 : Die Weddas von Ceylon und die sie umgebenden
Volkerschaften. 4° Wiesbaden 1892/3.

Lewis, J. P. Manual of the Vanni districts (Vavuniya and Mullailtivui of
the Northern Province, Ceylon. Colombo 1895.

Lawrie, a. C. a gazetteer of the Central Province of Ceylon, excluding
Walapane. 2 vols. Colombo 1 896 '8.

Schmidt, E. Ceylon. Berlin [1897].

Je\t:rs, R.W. Manual of the North-Central Province, Ceylon. Colombi) 1899.

Coomaraswajiy, a. K. Medixval Sinhalese art. Being a monograph on
medieval Sinhalese arts and crafts, mainly as surviving in the eigh-
teenth century, with an account of the structure of society and the
status of the craftsmen. 4° Norman Chapel, Broad Campden 1908.

Parker, H. Ancient Ceylon, an account of the aborigines and of i)art (jf
the early civilisation. London 1909.

Parker, H. Village folk-tales of Ceylon. Vol. I. London 1910.

Seligmann, C. G., and Z. Brenda. The Veddas. London 191 1.

CONTENTS.

Page

Introduction 1—9

SocuL Organisation 9—24

A. Historical 9

B. Descriptive 21

Castes and Caste-Groups 24— US

A. Special Groups 24

B. The Village Community 42

C. Subsidiary Professional Castes 85

D. Urban Castes 94

E. Nomadic Castes 'oo

F. Hill Tribes 112

G. Muslim Race Titles . . . .- '39

Appendices 146—172

A. Summary of Caste-Groups 146

B. Caste Index '53

Table of Languages '^

Table of Religions '7°

Table of Forest Tribes ■ '7'

List of the more important works on Indian Ethnography 173—211

CORRECTIONS.

p. 6, line 4 for belongs read belong. — P. 6, line 17 for stand read st.mds.

P. 13, line 31 for others read other. — P. 25, line 36 for between read about.

P. 27, line 5 for clouth read South. — P. 30, line 16 for times' read lime's.

P. 30, line 42 for Chapter read review.

BINDING SECT. FEB 5 1975

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY

DS Eaines, (Sir) Jerx-cise

430 Athelstane

B23 2thnograj:hy (castes and

tribes)